Blame view
fs/btrfs/scrub.c
110 KB
a2de733c7 btrfs: scrub |
1 |
/* |
b6bfebc13 Btrfs: cleanup sc... |
2 |
* Copyright (C) 2011, 2012 STRATO. All rights reserved. |
a2de733c7 btrfs: scrub |
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
* * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public * License v2 as published by the Free Software Foundation. * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * General Public License for more details. * * You should have received a copy of the GNU General Public * License along with this program; if not, write to the * Free Software Foundation, Inc., 59 Temple Place - Suite 330, * Boston, MA 021110-1307, USA. */ |
a2de733c7 btrfs: scrub |
18 |
#include <linux/blkdev.h> |
558540c17 btrfs scrub: prin... |
19 |
#include <linux/ratelimit.h> |
a2de733c7 btrfs: scrub |
20 21 22 23 |
#include "ctree.h" #include "volumes.h" #include "disk-io.h" #include "ordered-data.h" |
0ef8e4515 btrfs scrub: add ... |
24 |
#include "transaction.h" |
558540c17 btrfs scrub: prin... |
25 |
#include "backref.h" |
5da6fcbc4 btrfs: integratin... |
26 |
#include "extent_io.h" |
ff023aac3 Btrfs: add code t... |
27 |
#include "dev-replace.h" |
21adbd5cb Btrfs: integrate ... |
28 |
#include "check-integrity.h" |
606686eea Btrfs: use rcu to... |
29 |
#include "rcu-string.h" |
53b381b3a Btrfs: RAID5 and ... |
30 |
#include "raid56.h" |
a2de733c7 btrfs: scrub |
31 32 33 34 35 36 37 38 |
/* * This is only the first step towards a full-features scrub. It reads all * extent and super block and verifies the checksums. In case a bad checksum * is found or the extent cannot be read, good data will be written back if * any can be found. * * Future enhancements: |
a2de733c7 btrfs: scrub |
39 40 |
* - In case an unrepairable extent is encountered, track which files are * affected and report them |
a2de733c7 btrfs: scrub |
41 |
* - track and record media errors, throw out bad devices |
a2de733c7 btrfs: scrub |
42 |
* - add a mode to also read unallocated space |
a2de733c7 btrfs: scrub |
43 |
*/ |
b5d67f64f Btrfs: change scr... |
44 |
struct scrub_block; |
d9d181c1b Btrfs: rename the... |
45 |
struct scrub_ctx; |
a2de733c7 btrfs: scrub |
46 |
|
ff023aac3 Btrfs: add code t... |
47 48 49 50 51 52 53 54 55 |
/* * the following three values only influence the performance. * The last one configures the number of parallel and outstanding I/O * operations. The first two values configure an upper limit for the number * of (dynamically allocated) pages that are added to a bio. */ #define SCRUB_PAGES_PER_RD_BIO 32 /* 128k per bio */ #define SCRUB_PAGES_PER_WR_BIO 32 /* 128k per bio */ #define SCRUB_BIOS_PER_SCTX 64 /* 8MB per device in flight */ |
7a9e99876 Btrfs: make the s... |
56 57 58 59 60 61 |
/* * the following value times PAGE_SIZE needs to be large enough to match the * largest node/leaf/sector size that shall be supported. * Values larger than BTRFS_STRIPE_LEN are not supported. */ |
b5d67f64f Btrfs: change scr... |
62 |
#define SCRUB_MAX_PAGES_PER_BLOCK 16 /* 64k per node/leaf/sector */ |
a2de733c7 btrfs: scrub |
63 |
|
af8e2d1df Btrfs, scrub: rep... |
64 65 66 67 68 69 |
struct scrub_recover { atomic_t refs; struct btrfs_bio *bbio; u64 *raid_map; u64 map_length; }; |
a2de733c7 btrfs: scrub |
70 |
struct scrub_page { |
b5d67f64f Btrfs: change scr... |
71 72 |
struct scrub_block *sblock; struct page *page; |
442a4f630 Btrfs: add device... |
73 |
struct btrfs_device *dev; |
5a6ac9eac Btrfs, raid56: su... |
74 |
struct list_head list; |
a2de733c7 btrfs: scrub |
75 76 |
u64 flags; /* extent flags */ u64 generation; |
b5d67f64f Btrfs: change scr... |
77 78 |
u64 logical; u64 physical; |
ff023aac3 Btrfs: add code t... |
79 |
u64 physical_for_dev_replace; |
7a9e99876 Btrfs: make the s... |
80 |
atomic_t ref_count; |
b5d67f64f Btrfs: change scr... |
81 82 83 84 85 |
struct { unsigned int mirror_num:8; unsigned int have_csum:1; unsigned int io_error:1; }; |
a2de733c7 btrfs: scrub |
86 |
u8 csum[BTRFS_CSUM_SIZE]; |
af8e2d1df Btrfs, scrub: rep... |
87 88 |
struct scrub_recover *recover; |
a2de733c7 btrfs: scrub |
89 90 91 92 |
}; struct scrub_bio { int index; |
d9d181c1b Btrfs: rename the... |
93 |
struct scrub_ctx *sctx; |
a36cf8b89 Btrfs: remove the... |
94 |
struct btrfs_device *dev; |
a2de733c7 btrfs: scrub |
95 96 97 98 |
struct bio *bio; int err; u64 logical; u64 physical; |
ff023aac3 Btrfs: add code t... |
99 100 101 102 103 |
#if SCRUB_PAGES_PER_WR_BIO >= SCRUB_PAGES_PER_RD_BIO struct scrub_page *pagev[SCRUB_PAGES_PER_WR_BIO]; #else struct scrub_page *pagev[SCRUB_PAGES_PER_RD_BIO]; #endif |
b5d67f64f Btrfs: change scr... |
104 |
int page_count; |
a2de733c7 btrfs: scrub |
105 106 107 |
int next_free; struct btrfs_work work; }; |
b5d67f64f Btrfs: change scr... |
108 |
struct scrub_block { |
7a9e99876 Btrfs: make the s... |
109 |
struct scrub_page *pagev[SCRUB_MAX_PAGES_PER_BLOCK]; |
b5d67f64f Btrfs: change scr... |
110 111 112 |
int page_count; atomic_t outstanding_pages; atomic_t ref_count; /* free mem on transition to zero */ |
d9d181c1b Btrfs: rename the... |
113 |
struct scrub_ctx *sctx; |
5a6ac9eac Btrfs, raid56: su... |
114 |
struct scrub_parity *sparity; |
b5d67f64f Btrfs: change scr... |
115 116 117 118 |
struct { unsigned int header_error:1; unsigned int checksum_error:1; unsigned int no_io_error_seen:1; |
442a4f630 Btrfs: add device... |
119 |
unsigned int generation_error:1; /* also sets header_error */ |
5a6ac9eac Btrfs, raid56: su... |
120 121 122 123 |
/* The following is for the data used to check parity */ /* It is for the data with checksum */ unsigned int data_corrected:1; |
b5d67f64f Btrfs: change scr... |
124 125 |
}; }; |
5a6ac9eac Btrfs, raid56: su... |
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
/* Used for the chunks with parity stripe such RAID5/6 */ struct scrub_parity { struct scrub_ctx *sctx; struct btrfs_device *scrub_dev; u64 logic_start; u64 logic_end; int nsectors; int stripe_len; atomic_t ref_count; struct list_head spages; /* Work of parity check and repair */ struct btrfs_work work; /* Mark the parity blocks which have data */ unsigned long *dbitmap; /* * Mark the parity blocks which have data, but errors happen when * read data or check data */ unsigned long *ebitmap; unsigned long bitmap[0]; }; |
ff023aac3 Btrfs: add code t... |
158 159 160 161 162 163 164 |
struct scrub_wr_ctx { struct scrub_bio *wr_curr_bio; struct btrfs_device *tgtdev; int pages_per_wr_bio; /* <= SCRUB_PAGES_PER_WR_BIO */ atomic_t flush_all_writes; struct mutex wr_lock; }; |
d9d181c1b Btrfs: rename the... |
165 |
struct scrub_ctx { |
ff023aac3 Btrfs: add code t... |
166 |
struct scrub_bio *bios[SCRUB_BIOS_PER_SCTX]; |
a36cf8b89 Btrfs: remove the... |
167 |
struct btrfs_root *dev_root; |
a2de733c7 btrfs: scrub |
168 169 |
int first_free; int curr; |
b6bfebc13 Btrfs: cleanup sc... |
170 171 |
atomic_t bios_in_flight; atomic_t workers_pending; |
a2de733c7 btrfs: scrub |
172 173 174 175 176 |
spinlock_t list_lock; wait_queue_head_t list_wait; u16 csum_size; struct list_head csum_list; atomic_t cancel_req; |
8628764e1 btrfs: add readon... |
177 |
int readonly; |
ff023aac3 Btrfs: add code t... |
178 |
int pages_per_rd_bio; |
b5d67f64f Btrfs: change scr... |
179 180 |
u32 sectorsize; u32 nodesize; |
63a212abc Btrfs: disallow s... |
181 182 |
int is_dev_replace; |
ff023aac3 Btrfs: add code t... |
183 |
struct scrub_wr_ctx wr_ctx; |
63a212abc Btrfs: disallow s... |
184 |
|
a2de733c7 btrfs: scrub |
185 186 187 188 189 190 |
/* * statistics */ struct btrfs_scrub_progress stat; spinlock_t stat_lock; }; |
0ef8e4515 btrfs scrub: add ... |
191 |
struct scrub_fixup_nodatasum { |
d9d181c1b Btrfs: rename the... |
192 |
struct scrub_ctx *sctx; |
a36cf8b89 Btrfs: remove the... |
193 |
struct btrfs_device *dev; |
0ef8e4515 btrfs scrub: add ... |
194 195 196 197 198 |
u64 logical; struct btrfs_root *root; struct btrfs_work work; int mirror_num; }; |
652f25a29 Btrfs: improve re... |
199 200 201 202 203 204 |
struct scrub_nocow_inode { u64 inum; u64 offset; u64 root; struct list_head list; }; |
ff023aac3 Btrfs: add code t... |
205 206 207 208 209 210 |
struct scrub_copy_nocow_ctx { struct scrub_ctx *sctx; u64 logical; u64 len; int mirror_num; u64 physical_for_dev_replace; |
652f25a29 Btrfs: improve re... |
211 |
struct list_head inodes; |
ff023aac3 Btrfs: add code t... |
212 213 |
struct btrfs_work work; }; |
558540c17 btrfs scrub: prin... |
214 215 216 |
struct scrub_warning { struct btrfs_path *path; u64 extent_item_size; |
558540c17 btrfs scrub: prin... |
217 218 219 220 |
const char *errstr; sector_t sector; u64 logical; struct btrfs_device *dev; |
558540c17 btrfs scrub: prin... |
221 |
}; |
b6bfebc13 Btrfs: cleanup sc... |
222 223 224 225 |
static void scrub_pending_bio_inc(struct scrub_ctx *sctx); static void scrub_pending_bio_dec(struct scrub_ctx *sctx); static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx); static void scrub_pending_trans_workers_dec(struct scrub_ctx *sctx); |
b5d67f64f Btrfs: change scr... |
226 |
static int scrub_handle_errored_block(struct scrub_block *sblock_to_check); |
d9d181c1b Btrfs: rename the... |
227 |
static int scrub_setup_recheck_block(struct scrub_ctx *sctx, |
3ec706c83 Btrfs: pass fs_in... |
228 |
struct btrfs_fs_info *fs_info, |
ff023aac3 Btrfs: add code t... |
229 |
struct scrub_block *original_sblock, |
b5d67f64f Btrfs: change scr... |
230 |
u64 length, u64 logical, |
ff023aac3 Btrfs: add code t... |
231 |
struct scrub_block *sblocks_for_recheck); |
34f5c8e90 Btrfs: in scrub r... |
232 233 234 |
static void scrub_recheck_block(struct btrfs_fs_info *fs_info, struct scrub_block *sblock, int is_metadata, int have_csum, u8 *csum, u64 generation, |
af8e2d1df Btrfs, scrub: rep... |
235 |
u16 csum_size, int retry_failed_mirror); |
b5d67f64f Btrfs: change scr... |
236 237 238 239 240 |
static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info, struct scrub_block *sblock, int is_metadata, int have_csum, const u8 *csum, u64 generation, u16 csum_size); |
b5d67f64f Btrfs: change scr... |
241 242 243 244 245 246 |
static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int force_write); static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int page_num, int force_write); |
ff023aac3 Btrfs: add code t... |
247 248 249 |
static void scrub_write_block_to_dev_replace(struct scrub_block *sblock); static int scrub_write_page_to_dev_replace(struct scrub_block *sblock, int page_num); |
b5d67f64f Btrfs: change scr... |
250 251 252 253 254 |
static int scrub_checksum_data(struct scrub_block *sblock); static int scrub_checksum_tree_block(struct scrub_block *sblock); static int scrub_checksum_super(struct scrub_block *sblock); static void scrub_block_get(struct scrub_block *sblock); static void scrub_block_put(struct scrub_block *sblock); |
7a9e99876 Btrfs: make the s... |
255 256 |
static void scrub_page_get(struct scrub_page *spage); static void scrub_page_put(struct scrub_page *spage); |
5a6ac9eac Btrfs, raid56: su... |
257 258 |
static void scrub_parity_get(struct scrub_parity *sparity); static void scrub_parity_put(struct scrub_parity *sparity); |
ff023aac3 Btrfs: add code t... |
259 260 |
static int scrub_add_page_to_rd_bio(struct scrub_ctx *sctx, struct scrub_page *spage); |
d9d181c1b Btrfs: rename the... |
261 |
static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len, |
a36cf8b89 Btrfs: remove the... |
262 |
u64 physical, struct btrfs_device *dev, u64 flags, |
ff023aac3 Btrfs: add code t... |
263 264 |
u64 gen, int mirror_num, u8 *csum, int force, u64 physical_for_dev_replace); |
1623edebe Btrfs: minor clea... |
265 |
static void scrub_bio_end_io(struct bio *bio, int err); |
b5d67f64f Btrfs: change scr... |
266 267 |
static void scrub_bio_end_io_worker(struct btrfs_work *work); static void scrub_block_complete(struct scrub_block *sblock); |
ff023aac3 Btrfs: add code t... |
268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 |
static void scrub_remap_extent(struct btrfs_fs_info *fs_info, u64 extent_logical, u64 extent_len, u64 *extent_physical, struct btrfs_device **extent_dev, int *extent_mirror_num); static int scrub_setup_wr_ctx(struct scrub_ctx *sctx, struct scrub_wr_ctx *wr_ctx, struct btrfs_fs_info *fs_info, struct btrfs_device *dev, int is_dev_replace); static void scrub_free_wr_ctx(struct scrub_wr_ctx *wr_ctx); static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx, struct scrub_page *spage); static void scrub_wr_submit(struct scrub_ctx *sctx); static void scrub_wr_bio_end_io(struct bio *bio, int err); static void scrub_wr_bio_end_io_worker(struct btrfs_work *work); static int write_page_nocow(struct scrub_ctx *sctx, u64 physical_for_dev_replace, struct page *page); static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root, |
652f25a29 Btrfs: improve re... |
287 |
struct scrub_copy_nocow_ctx *ctx); |
ff023aac3 Btrfs: add code t... |
288 289 290 |
static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len, int mirror_num, u64 physical_for_dev_replace); static void copy_nocow_pages_worker(struct btrfs_work *work); |
cb7ab0215 Btrfs: wrap repea... |
291 |
static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info); |
3cb0929ad Btrfs: fix wrong ... |
292 |
static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info); |
1623edebe Btrfs: minor clea... |
293 |
|
b6bfebc13 Btrfs: cleanup sc... |
294 295 296 297 298 299 300 301 302 303 |
static void scrub_pending_bio_inc(struct scrub_ctx *sctx) { atomic_inc(&sctx->bios_in_flight); } static void scrub_pending_bio_dec(struct scrub_ctx *sctx) { atomic_dec(&sctx->bios_in_flight); wake_up(&sctx->list_wait); } |
cb7ab0215 Btrfs: wrap repea... |
304 |
static void __scrub_blocked_if_needed(struct btrfs_fs_info *fs_info) |
3cb0929ad Btrfs: fix wrong ... |
305 306 307 308 309 310 311 312 |
{ while (atomic_read(&fs_info->scrub_pause_req)) { mutex_unlock(&fs_info->scrub_lock); wait_event(fs_info->scrub_pause_wait, atomic_read(&fs_info->scrub_pause_req) == 0); mutex_lock(&fs_info->scrub_lock); } } |
cb7ab0215 Btrfs: wrap repea... |
313 314 315 316 317 318 319 320 321 322 323 324 |
static void scrub_blocked_if_needed(struct btrfs_fs_info *fs_info) { atomic_inc(&fs_info->scrubs_paused); wake_up(&fs_info->scrub_pause_wait); mutex_lock(&fs_info->scrub_lock); __scrub_blocked_if_needed(fs_info); atomic_dec(&fs_info->scrubs_paused); mutex_unlock(&fs_info->scrub_lock); wake_up(&fs_info->scrub_pause_wait); } |
b6bfebc13 Btrfs: cleanup sc... |
325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 |
/* * used for workers that require transaction commits (i.e., for the * NOCOW case) */ static void scrub_pending_trans_workers_inc(struct scrub_ctx *sctx) { struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; /* * increment scrubs_running to prevent cancel requests from * completing as long as a worker is running. we must also * increment scrubs_paused to prevent deadlocking on pause * requests used for transactions commits (as the worker uses a * transaction context). it is safe to regard the worker * as paused for all matters practical. effectively, we only * avoid cancellation requests from completing. */ mutex_lock(&fs_info->scrub_lock); atomic_inc(&fs_info->scrubs_running); atomic_inc(&fs_info->scrubs_paused); mutex_unlock(&fs_info->scrub_lock); |
32a447896 Btrfs: wake up @s... |
346 347 348 349 350 351 352 353 354 |
/* * check if @scrubs_running=@scrubs_paused condition * inside wait_event() is not an atomic operation. * which means we may inc/dec @scrub_running/paused * at any time. Let's wake up @scrub_pause_wait as * much as we can to let commit transaction blocked less. */ wake_up(&fs_info->scrub_pause_wait); |
b6bfebc13 Btrfs: cleanup sc... |
355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 |
atomic_inc(&sctx->workers_pending); } /* used for workers that require transaction commits */ static void scrub_pending_trans_workers_dec(struct scrub_ctx *sctx) { struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; /* * see scrub_pending_trans_workers_inc() why we're pretending * to be paused in the scrub counters */ mutex_lock(&fs_info->scrub_lock); atomic_dec(&fs_info->scrubs_running); atomic_dec(&fs_info->scrubs_paused); mutex_unlock(&fs_info->scrub_lock); atomic_dec(&sctx->workers_pending); wake_up(&fs_info->scrub_pause_wait); wake_up(&sctx->list_wait); } |
d9d181c1b Btrfs: rename the... |
375 |
static void scrub_free_csums(struct scrub_ctx *sctx) |
a2de733c7 btrfs: scrub |
376 |
{ |
d9d181c1b Btrfs: rename the... |
377 |
while (!list_empty(&sctx->csum_list)) { |
a2de733c7 btrfs: scrub |
378 |
struct btrfs_ordered_sum *sum; |
d9d181c1b Btrfs: rename the... |
379 |
sum = list_first_entry(&sctx->csum_list, |
a2de733c7 btrfs: scrub |
380 381 382 383 384 |
struct btrfs_ordered_sum, list); list_del(&sum->list); kfree(sum); } } |
d9d181c1b Btrfs: rename the... |
385 |
static noinline_for_stack void scrub_free_ctx(struct scrub_ctx *sctx) |
a2de733c7 btrfs: scrub |
386 387 |
{ int i; |
a2de733c7 btrfs: scrub |
388 |
|
d9d181c1b Btrfs: rename the... |
389 |
if (!sctx) |
a2de733c7 btrfs: scrub |
390 |
return; |
ff023aac3 Btrfs: add code t... |
391 |
scrub_free_wr_ctx(&sctx->wr_ctx); |
b5d67f64f Btrfs: change scr... |
392 |
/* this can happen when scrub is cancelled */ |
d9d181c1b Btrfs: rename the... |
393 394 |
if (sctx->curr != -1) { struct scrub_bio *sbio = sctx->bios[sctx->curr]; |
b5d67f64f Btrfs: change scr... |
395 396 |
for (i = 0; i < sbio->page_count; i++) { |
ff023aac3 Btrfs: add code t... |
397 |
WARN_ON(!sbio->pagev[i]->page); |
b5d67f64f Btrfs: change scr... |
398 399 400 401 |
scrub_block_put(sbio->pagev[i]->sblock); } bio_put(sbio->bio); } |
ff023aac3 Btrfs: add code t... |
402 |
for (i = 0; i < SCRUB_BIOS_PER_SCTX; ++i) { |
d9d181c1b Btrfs: rename the... |
403 |
struct scrub_bio *sbio = sctx->bios[i]; |
a2de733c7 btrfs: scrub |
404 405 406 |
if (!sbio) break; |
a2de733c7 btrfs: scrub |
407 408 |
kfree(sbio); } |
d9d181c1b Btrfs: rename the... |
409 410 |
scrub_free_csums(sctx); kfree(sctx); |
a2de733c7 btrfs: scrub |
411 412 413 |
} static noinline_for_stack |
63a212abc Btrfs: disallow s... |
414 |
struct scrub_ctx *scrub_setup_ctx(struct btrfs_device *dev, int is_dev_replace) |
a2de733c7 btrfs: scrub |
415 |
{ |
d9d181c1b Btrfs: rename the... |
416 |
struct scrub_ctx *sctx; |
a2de733c7 btrfs: scrub |
417 |
int i; |
a2de733c7 btrfs: scrub |
418 |
struct btrfs_fs_info *fs_info = dev->dev_root->fs_info; |
ff023aac3 Btrfs: add code t... |
419 420 |
int pages_per_rd_bio; int ret; |
a2de733c7 btrfs: scrub |
421 |
|
ff023aac3 Btrfs: add code t... |
422 423 424 425 426 427 428 429 430 431 432 433 |
/* * the setting of pages_per_rd_bio is correct for scrub but might * be wrong for the dev_replace code where we might read from * different devices in the initial huge bios. However, that * code is able to correctly handle the case when adding a page * to a bio fails. */ if (dev->bdev) pages_per_rd_bio = min_t(int, SCRUB_PAGES_PER_RD_BIO, bio_get_nr_vecs(dev->bdev)); else pages_per_rd_bio = SCRUB_PAGES_PER_RD_BIO; |
d9d181c1b Btrfs: rename the... |
434 435 |
sctx = kzalloc(sizeof(*sctx), GFP_NOFS); if (!sctx) |
a2de733c7 btrfs: scrub |
436 |
goto nomem; |
63a212abc Btrfs: disallow s... |
437 |
sctx->is_dev_replace = is_dev_replace; |
ff023aac3 Btrfs: add code t... |
438 |
sctx->pages_per_rd_bio = pages_per_rd_bio; |
d9d181c1b Btrfs: rename the... |
439 |
sctx->curr = -1; |
a36cf8b89 Btrfs: remove the... |
440 |
sctx->dev_root = dev->dev_root; |
ff023aac3 Btrfs: add code t... |
441 |
for (i = 0; i < SCRUB_BIOS_PER_SCTX; ++i) { |
a2de733c7 btrfs: scrub |
442 443 444 445 446 |
struct scrub_bio *sbio; sbio = kzalloc(sizeof(*sbio), GFP_NOFS); if (!sbio) goto nomem; |
d9d181c1b Btrfs: rename the... |
447 |
sctx->bios[i] = sbio; |
a2de733c7 btrfs: scrub |
448 |
|
a2de733c7 btrfs: scrub |
449 |
sbio->index = i; |
d9d181c1b Btrfs: rename the... |
450 |
sbio->sctx = sctx; |
b5d67f64f Btrfs: change scr... |
451 |
sbio->page_count = 0; |
9e0af2376 Btrfs: fix task h... |
452 453 |
btrfs_init_work(&sbio->work, btrfs_scrub_helper, scrub_bio_end_io_worker, NULL, NULL); |
a2de733c7 btrfs: scrub |
454 |
|
ff023aac3 Btrfs: add code t... |
455 |
if (i != SCRUB_BIOS_PER_SCTX - 1) |
d9d181c1b Btrfs: rename the... |
456 |
sctx->bios[i]->next_free = i + 1; |
0ef8e4515 btrfs scrub: add ... |
457 |
else |
d9d181c1b Btrfs: rename the... |
458 459 460 461 |
sctx->bios[i]->next_free = -1; } sctx->first_free = 0; sctx->nodesize = dev->dev_root->nodesize; |
d9d181c1b Btrfs: rename the... |
462 |
sctx->sectorsize = dev->dev_root->sectorsize; |
b6bfebc13 Btrfs: cleanup sc... |
463 464 |
atomic_set(&sctx->bios_in_flight, 0); atomic_set(&sctx->workers_pending, 0); |
d9d181c1b Btrfs: rename the... |
465 466 467 468 469 470 471 |
atomic_set(&sctx->cancel_req, 0); sctx->csum_size = btrfs_super_csum_size(fs_info->super_copy); INIT_LIST_HEAD(&sctx->csum_list); spin_lock_init(&sctx->list_lock); spin_lock_init(&sctx->stat_lock); init_waitqueue_head(&sctx->list_wait); |
ff023aac3 Btrfs: add code t... |
472 473 474 475 476 477 478 |
ret = scrub_setup_wr_ctx(sctx, &sctx->wr_ctx, fs_info, fs_info->dev_replace.tgtdev, is_dev_replace); if (ret) { scrub_free_ctx(sctx); return ERR_PTR(ret); } |
d9d181c1b Btrfs: rename the... |
479 |
return sctx; |
a2de733c7 btrfs: scrub |
480 481 |
nomem: |
d9d181c1b Btrfs: rename the... |
482 |
scrub_free_ctx(sctx); |
a2de733c7 btrfs: scrub |
483 484 |
return ERR_PTR(-ENOMEM); } |
ff023aac3 Btrfs: add code t... |
485 486 |
static int scrub_print_warning_inode(u64 inum, u64 offset, u64 root, void *warn_ctx) |
558540c17 btrfs scrub: prin... |
487 488 489 490 491 492 493 |
{ u64 isize; u32 nlink; int ret; int i; struct extent_buffer *eb; struct btrfs_inode_item *inode_item; |
ff023aac3 Btrfs: add code t... |
494 |
struct scrub_warning *swarn = warn_ctx; |
558540c17 btrfs scrub: prin... |
495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 |
struct btrfs_fs_info *fs_info = swarn->dev->dev_root->fs_info; struct inode_fs_paths *ipath = NULL; struct btrfs_root *local_root; struct btrfs_key root_key; root_key.objectid = root; root_key.type = BTRFS_ROOT_ITEM_KEY; root_key.offset = (u64)-1; local_root = btrfs_read_fs_root_no_name(fs_info, &root_key); if (IS_ERR(local_root)) { ret = PTR_ERR(local_root); goto err; } ret = inode_item_info(inum, 0, local_root, swarn->path); if (ret) { btrfs_release_path(swarn->path); goto err; } eb = swarn->path->nodes[0]; inode_item = btrfs_item_ptr(eb, swarn->path->slots[0], struct btrfs_inode_item); isize = btrfs_inode_size(eb, inode_item); nlink = btrfs_inode_nlink(eb, inode_item); btrfs_release_path(swarn->path); ipath = init_ipath(4096, local_root, swarn->path); |
26bdef541 btrfs scrub: hand... |
523 524 525 526 527 |
if (IS_ERR(ipath)) { ret = PTR_ERR(ipath); ipath = NULL; goto err; } |
558540c17 btrfs scrub: prin... |
528 529 530 531 532 533 534 535 536 537 |
ret = paths_from_inode(inum, ipath); if (ret < 0) goto err; /* * we deliberately ignore the bit ipath might have been too small to * hold all of the paths here */ for (i = 0; i < ipath->fspath->elem_cnt; ++i) |
efe120a06 Btrfs: convert pr... |
538 |
printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev " |
558540c17 btrfs scrub: prin... |
539 540 541 |
"%s, sector %llu, root %llu, inode %llu, offset %llu, " "length %llu, links %u (path: %s) ", swarn->errstr, |
606686eea Btrfs: use rcu to... |
542 |
swarn->logical, rcu_str_deref(swarn->dev->name), |
558540c17 btrfs scrub: prin... |
543 544 |
(unsigned long long)swarn->sector, root, inum, offset, min(isize - offset, (u64)PAGE_SIZE), nlink, |
745c4d8e1 btrfs: Fix up 32/... |
545 |
(char *)(unsigned long)ipath->fspath->val[i]); |
558540c17 btrfs scrub: prin... |
546 547 548 549 550 |
free_ipath(ipath); return 0; err: |
efe120a06 Btrfs: convert pr... |
551 |
printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev " |
558540c17 btrfs scrub: prin... |
552 553 554 |
"%s, sector %llu, root %llu, inode %llu, offset %llu: path " "resolving failed with ret=%d ", swarn->errstr, |
606686eea Btrfs: use rcu to... |
555 |
swarn->logical, rcu_str_deref(swarn->dev->name), |
558540c17 btrfs scrub: prin... |
556 557 558 559 560 |
(unsigned long long)swarn->sector, root, inum, offset, ret); free_ipath(ipath); return 0; } |
b5d67f64f Btrfs: change scr... |
561 |
static void scrub_print_warning(const char *errstr, struct scrub_block *sblock) |
558540c17 btrfs scrub: prin... |
562 |
{ |
a36cf8b89 Btrfs: remove the... |
563 564 |
struct btrfs_device *dev; struct btrfs_fs_info *fs_info; |
558540c17 btrfs scrub: prin... |
565 566 567 568 569 |
struct btrfs_path *path; struct btrfs_key found_key; struct extent_buffer *eb; struct btrfs_extent_item *ei; struct scrub_warning swarn; |
69917e431 Btrfs: fix a bug ... |
570 571 572 |
unsigned long ptr = 0; u64 extent_item_pos; u64 flags = 0; |
558540c17 btrfs scrub: prin... |
573 |
u64 ref_root; |
69917e431 Btrfs: fix a bug ... |
574 |
u32 item_size; |
558540c17 btrfs scrub: prin... |
575 |
u8 ref_level; |
69917e431 Btrfs: fix a bug ... |
576 |
int ret; |
558540c17 btrfs scrub: prin... |
577 |
|
a36cf8b89 Btrfs: remove the... |
578 |
WARN_ON(sblock->page_count < 1); |
7a9e99876 Btrfs: make the s... |
579 |
dev = sblock->pagev[0]->dev; |
a36cf8b89 Btrfs: remove the... |
580 |
fs_info = sblock->sctx->dev_root->fs_info; |
558540c17 btrfs scrub: prin... |
581 |
path = btrfs_alloc_path(); |
8b9456da0 btrfs: remove unu... |
582 583 |
if (!path) return; |
558540c17 btrfs scrub: prin... |
584 |
|
7a9e99876 Btrfs: make the s... |
585 586 |
swarn.sector = (sblock->pagev[0]->physical) >> 9; swarn.logical = sblock->pagev[0]->logical; |
558540c17 btrfs scrub: prin... |
587 |
swarn.errstr = errstr; |
a36cf8b89 Btrfs: remove the... |
588 |
swarn.dev = NULL; |
558540c17 btrfs scrub: prin... |
589 |
|
69917e431 Btrfs: fix a bug ... |
590 591 |
ret = extent_from_logical(fs_info, swarn.logical, path, &found_key, &flags); |
558540c17 btrfs scrub: prin... |
592 593 |
if (ret < 0) goto out; |
4692cf58a Btrfs: new backre... |
594 |
extent_item_pos = swarn.logical - found_key.objectid; |
558540c17 btrfs scrub: prin... |
595 596 597 598 599 |
swarn.extent_item_size = found_key.offset; eb = path->nodes[0]; ei = btrfs_item_ptr(eb, path->slots[0], struct btrfs_extent_item); item_size = btrfs_item_size_nr(eb, path->slots[0]); |
69917e431 Btrfs: fix a bug ... |
600 |
if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { |
558540c17 btrfs scrub: prin... |
601 |
do { |
6eda71d0c Btrfs: fix scrub_... |
602 603 604 |
ret = tree_backref_for_extent(&ptr, eb, &found_key, ei, item_size, &ref_root, &ref_level); |
606686eea Btrfs: use rcu to... |
605 |
printk_in_rcu(KERN_WARNING |
efe120a06 Btrfs: convert pr... |
606 |
"BTRFS: %s at logical %llu on dev %s, " |
558540c17 btrfs scrub: prin... |
607 |
"sector %llu: metadata %s (level %d) in tree " |
606686eea Btrfs: use rcu to... |
608 609 610 |
"%llu ", errstr, swarn.logical, rcu_str_deref(dev->name), |
558540c17 btrfs scrub: prin... |
611 612 613 614 615 |
(unsigned long long)swarn.sector, ref_level ? "node" : "leaf", ret < 0 ? -1 : ref_level, ret < 0 ? -1 : ref_root); } while (ret != 1); |
d8fe29e9d Btrfs: don't drop... |
616 |
btrfs_release_path(path); |
558540c17 btrfs scrub: prin... |
617 |
} else { |
d8fe29e9d Btrfs: don't drop... |
618 |
btrfs_release_path(path); |
558540c17 btrfs scrub: prin... |
619 |
swarn.path = path; |
a36cf8b89 Btrfs: remove the... |
620 |
swarn.dev = dev; |
7a3ae2f8c Btrfs: fix regres... |
621 622 |
iterate_extent_inodes(fs_info, found_key.objectid, extent_item_pos, 1, |
558540c17 btrfs scrub: prin... |
623 624 625 626 627 |
scrub_print_warning_inode, &swarn); } out: btrfs_free_path(path); |
558540c17 btrfs scrub: prin... |
628 |
} |
ff023aac3 Btrfs: add code t... |
629 |
static int scrub_fixup_readpage(u64 inum, u64 offset, u64 root, void *fixup_ctx) |
0ef8e4515 btrfs scrub: add ... |
630 |
{ |
5da6fcbc4 btrfs: integratin... |
631 |
struct page *page = NULL; |
0ef8e4515 btrfs scrub: add ... |
632 |
unsigned long index; |
ff023aac3 Btrfs: add code t... |
633 |
struct scrub_fixup_nodatasum *fixup = fixup_ctx; |
0ef8e4515 btrfs scrub: add ... |
634 |
int ret; |
5da6fcbc4 btrfs: integratin... |
635 |
int corrected = 0; |
0ef8e4515 btrfs scrub: add ... |
636 |
struct btrfs_key key; |
5da6fcbc4 btrfs: integratin... |
637 |
struct inode *inode = NULL; |
6f1c36055 Btrfs: fix race b... |
638 |
struct btrfs_fs_info *fs_info; |
0ef8e4515 btrfs scrub: add ... |
639 640 |
u64 end = offset + PAGE_SIZE - 1; struct btrfs_root *local_root; |
6f1c36055 Btrfs: fix race b... |
641 |
int srcu_index; |
0ef8e4515 btrfs scrub: add ... |
642 643 644 645 |
key.objectid = root; key.type = BTRFS_ROOT_ITEM_KEY; key.offset = (u64)-1; |
6f1c36055 Btrfs: fix race b... |
646 647 648 649 650 651 652 |
fs_info = fixup->root->fs_info; srcu_index = srcu_read_lock(&fs_info->subvol_srcu); local_root = btrfs_read_fs_root_no_name(fs_info, &key); if (IS_ERR(local_root)) { srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); |
0ef8e4515 btrfs scrub: add ... |
653 |
return PTR_ERR(local_root); |
6f1c36055 Btrfs: fix race b... |
654 |
} |
0ef8e4515 btrfs scrub: add ... |
655 656 657 658 |
key.type = BTRFS_INODE_ITEM_KEY; key.objectid = inum; key.offset = 0; |
6f1c36055 Btrfs: fix race b... |
659 660 |
inode = btrfs_iget(fs_info->sb, &key, local_root, NULL); srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); |
0ef8e4515 btrfs scrub: add ... |
661 662 |
if (IS_ERR(inode)) return PTR_ERR(inode); |
0ef8e4515 btrfs scrub: add ... |
663 664 665 |
index = offset >> PAGE_CACHE_SHIFT; page = find_or_create_page(inode->i_mapping, index, GFP_NOFS); |
5da6fcbc4 btrfs: integratin... |
666 667 668 669 670 671 |
if (!page) { ret = -ENOMEM; goto out; } if (PageUptodate(page)) { |
5da6fcbc4 btrfs: integratin... |
672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 |
if (PageDirty(page)) { /* * we need to write the data to the defect sector. the * data that was in that sector is not in memory, * because the page was modified. we must not write the * modified page to that sector. * * TODO: what could be done here: wait for the delalloc * runner to write out that page (might involve * COW) and see whether the sector is still * referenced afterwards. * * For the meantime, we'll treat this error * incorrectable, although there is a chance that a * later scrub will find the bad sector again and that * there's no dirty page in memory, then. */ ret = -EIO; goto out; } |
1203b6813 Btrfs: modify cle... |
692 |
ret = repair_io_failure(inode, offset, PAGE_SIZE, |
5da6fcbc4 btrfs: integratin... |
693 |
fixup->logical, page, |
ffdd2018d Btrfs: modify rep... |
694 |
offset - page_offset(page), |
5da6fcbc4 btrfs: integratin... |
695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 |
fixup->mirror_num); unlock_page(page); corrected = !ret; } else { /* * we need to get good data first. the general readpage path * will call repair_io_failure for us, we just have to make * sure we read the bad mirror. */ ret = set_extent_bits(&BTRFS_I(inode)->io_tree, offset, end, EXTENT_DAMAGED, GFP_NOFS); if (ret) { /* set_extent_bits should give proper error */ WARN_ON(ret > 0); if (ret > 0) ret = -EFAULT; goto out; } ret = extent_read_full_page(&BTRFS_I(inode)->io_tree, page, btrfs_get_extent, fixup->mirror_num); wait_on_page_locked(page); corrected = !test_range_bit(&BTRFS_I(inode)->io_tree, offset, end, EXTENT_DAMAGED, 0, NULL); if (!corrected) clear_extent_bits(&BTRFS_I(inode)->io_tree, offset, end, EXTENT_DAMAGED, GFP_NOFS); } out: if (page) put_page(page); |
7fb18a066 btrfs: Remove unn... |
729 730 |
iput(inode); |
0ef8e4515 btrfs scrub: add ... |
731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 |
if (ret < 0) return ret; if (ret == 0 && corrected) { /* * we only need to call readpage for one of the inodes belonging * to this extent. so make iterate_extent_inodes stop */ return 1; } return -EIO; } static void scrub_fixup_nodatasum(struct btrfs_work *work) { int ret; struct scrub_fixup_nodatasum *fixup; |
d9d181c1b Btrfs: rename the... |
750 |
struct scrub_ctx *sctx; |
0ef8e4515 btrfs scrub: add ... |
751 |
struct btrfs_trans_handle *trans = NULL; |
0ef8e4515 btrfs scrub: add ... |
752 753 754 755 |
struct btrfs_path *path; int uncorrectable = 0; fixup = container_of(work, struct scrub_fixup_nodatasum, work); |
d9d181c1b Btrfs: rename the... |
756 |
sctx = fixup->sctx; |
0ef8e4515 btrfs scrub: add ... |
757 758 759 |
path = btrfs_alloc_path(); if (!path) { |
d9d181c1b Btrfs: rename the... |
760 761 762 |
spin_lock(&sctx->stat_lock); ++sctx->stat.malloc_errors; spin_unlock(&sctx->stat_lock); |
0ef8e4515 btrfs scrub: add ... |
763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 |
uncorrectable = 1; goto out; } trans = btrfs_join_transaction(fixup->root); if (IS_ERR(trans)) { uncorrectable = 1; goto out; } /* * the idea is to trigger a regular read through the standard path. we * read a page from the (failed) logical address by specifying the * corresponding copynum of the failed sector. thus, that readpage is * expected to fail. * that is the point where on-the-fly error correction will kick in * (once it's finished) and rewrite the failed sector if a good copy * can be found. */ ret = iterate_inodes_from_logical(fixup->logical, fixup->root->fs_info, path, scrub_fixup_readpage, fixup); if (ret < 0) { uncorrectable = 1; goto out; } WARN_ON(ret != 1); |
d9d181c1b Btrfs: rename the... |
790 791 792 |
spin_lock(&sctx->stat_lock); ++sctx->stat.corrected_errors; spin_unlock(&sctx->stat_lock); |
0ef8e4515 btrfs scrub: add ... |
793 794 795 796 797 |
out: if (trans && !IS_ERR(trans)) btrfs_end_transaction(trans, fixup->root); if (uncorrectable) { |
d9d181c1b Btrfs: rename the... |
798 799 800 |
spin_lock(&sctx->stat_lock); ++sctx->stat.uncorrectable_errors; spin_unlock(&sctx->stat_lock); |
ff023aac3 Btrfs: add code t... |
801 802 803 |
btrfs_dev_replace_stats_inc( &sctx->dev_root->fs_info->dev_replace. num_uncorrectable_read_errors); |
efe120a06 Btrfs: convert pr... |
804 805 806 |
printk_ratelimited_in_rcu(KERN_ERR "BTRFS: " "unable to fixup (nodatasum) error at logical %llu on dev %s ", |
c1c9ff7c9 Btrfs: Remove sup... |
807 |
fixup->logical, rcu_str_deref(fixup->dev->name)); |
0ef8e4515 btrfs scrub: add ... |
808 809 810 811 |
} btrfs_free_path(path); kfree(fixup); |
b6bfebc13 Btrfs: cleanup sc... |
812 |
scrub_pending_trans_workers_dec(sctx); |
0ef8e4515 btrfs scrub: add ... |
813 |
} |
af8e2d1df Btrfs, scrub: rep... |
814 815 816 817 818 819 820 821 822 823 824 825 826 |
static inline void scrub_get_recover(struct scrub_recover *recover) { atomic_inc(&recover->refs); } static inline void scrub_put_recover(struct scrub_recover *recover) { if (atomic_dec_and_test(&recover->refs)) { kfree(recover->bbio); kfree(recover->raid_map); kfree(recover); } } |
a2de733c7 btrfs: scrub |
827 |
/* |
b5d67f64f Btrfs: change scr... |
828 829 830 831 832 833 |
* scrub_handle_errored_block gets called when either verification of the * pages failed or the bio failed to read, e.g. with EIO. In the latter * case, this function handles all pages in the bio, even though only one * may be bad. * The goal of this function is to repair the errored block by using the * contents of one of the mirrors. |
a2de733c7 btrfs: scrub |
834 |
*/ |
b5d67f64f Btrfs: change scr... |
835 |
static int scrub_handle_errored_block(struct scrub_block *sblock_to_check) |
a2de733c7 btrfs: scrub |
836 |
{ |
d9d181c1b Btrfs: rename the... |
837 |
struct scrub_ctx *sctx = sblock_to_check->sctx; |
a36cf8b89 Btrfs: remove the... |
838 |
struct btrfs_device *dev; |
b5d67f64f Btrfs: change scr... |
839 840 841 842 843 844 845 846 847 848 849 850 851 852 |
struct btrfs_fs_info *fs_info; u64 length; u64 logical; u64 generation; unsigned int failed_mirror_index; unsigned int is_metadata; unsigned int have_csum; u8 *csum; struct scrub_block *sblocks_for_recheck; /* holds one for each mirror */ struct scrub_block *sblock_bad; int ret; int mirror_index; int page_num; int success; |
558540c17 btrfs scrub: prin... |
853 |
static DEFINE_RATELIMIT_STATE(_rs, DEFAULT_RATELIMIT_INTERVAL, |
b5d67f64f Btrfs: change scr... |
854 855 856 |
DEFAULT_RATELIMIT_BURST); BUG_ON(sblock_to_check->page_count < 1); |
a36cf8b89 Btrfs: remove the... |
857 |
fs_info = sctx->dev_root->fs_info; |
4ded4f639 Btrfs: fix BUG() ... |
858 859 860 861 862 863 864 865 866 867 868 |
if (sblock_to_check->pagev[0]->flags & BTRFS_EXTENT_FLAG_SUPER) { /* * if we find an error in a super block, we just report it. * They will get written with the next transaction commit * anyway */ spin_lock(&sctx->stat_lock); ++sctx->stat.super_errors; spin_unlock(&sctx->stat_lock); return 0; } |
b5d67f64f Btrfs: change scr... |
869 |
length = sblock_to_check->page_count * PAGE_SIZE; |
7a9e99876 Btrfs: make the s... |
870 871 872 873 874 |
logical = sblock_to_check->pagev[0]->logical; generation = sblock_to_check->pagev[0]->generation; BUG_ON(sblock_to_check->pagev[0]->mirror_num < 1); failed_mirror_index = sblock_to_check->pagev[0]->mirror_num - 1; is_metadata = !(sblock_to_check->pagev[0]->flags & |
b5d67f64f Btrfs: change scr... |
875 |
BTRFS_EXTENT_FLAG_DATA); |
7a9e99876 Btrfs: make the s... |
876 877 878 |
have_csum = sblock_to_check->pagev[0]->have_csum; csum = sblock_to_check->pagev[0]->csum; dev = sblock_to_check->pagev[0]->dev; |
13db62b7a btrfs scrub: adde... |
879 |
|
ff023aac3 Btrfs: add code t... |
880 881 882 883 |
if (sctx->is_dev_replace && !is_metadata && !have_csum) { sblocks_for_recheck = NULL; goto nodatasum_case; } |
b5d67f64f Btrfs: change scr... |
884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 |
/* * read all mirrors one after the other. This includes to * re-read the extent or metadata block that failed (that was * the cause that this fixup code is called) another time, * page by page this time in order to know which pages * caused I/O errors and which ones are good (for all mirrors). * It is the goal to handle the situation when more than one * mirror contains I/O errors, but the errors do not * overlap, i.e. the data can be repaired by selecting the * pages from those mirrors without I/O error on the * particular pages. One example (with blocks >= 2 * PAGE_SIZE) * would be that mirror #1 has an I/O error on the first page, * the second page is good, and mirror #2 has an I/O error on * the second page, but the first page is good. * Then the first page of the first mirror can be repaired by * taking the first page of the second mirror, and the * second page of the second mirror can be repaired by * copying the contents of the 2nd page of the 1st mirror. * One more note: if the pages of one mirror contain I/O * errors, the checksum cannot be verified. In order to get * the best data for repairing, the first attempt is to find * a mirror without I/O errors and with a validated checksum. * Only if this is not possible, the pages are picked from * mirrors with I/O errors without considering the checksum. * If the latter is the case, at the end, the checksum of the * repaired area is verified in order to correctly maintain * the statistics. */ sblocks_for_recheck = kzalloc(BTRFS_MAX_MIRRORS * sizeof(*sblocks_for_recheck), GFP_NOFS); if (!sblocks_for_recheck) { |
d9d181c1b Btrfs: rename the... |
917 918 919 920 921 |
spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; sctx->stat.read_errors++; sctx->stat.uncorrectable_errors++; spin_unlock(&sctx->stat_lock); |
a36cf8b89 Btrfs: remove the... |
922 |
btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS); |
b5d67f64f Btrfs: change scr... |
923 |
goto out; |
a2de733c7 btrfs: scrub |
924 |
} |
b5d67f64f Btrfs: change scr... |
925 |
/* setup the context, map the logical blocks and alloc the pages */ |
ff023aac3 Btrfs: add code t... |
926 |
ret = scrub_setup_recheck_block(sctx, fs_info, sblock_to_check, length, |
b5d67f64f Btrfs: change scr... |
927 928 |
logical, sblocks_for_recheck); if (ret) { |
d9d181c1b Btrfs: rename the... |
929 930 931 932 |
spin_lock(&sctx->stat_lock); sctx->stat.read_errors++; sctx->stat.uncorrectable_errors++; spin_unlock(&sctx->stat_lock); |
a36cf8b89 Btrfs: remove the... |
933 |
btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS); |
b5d67f64f Btrfs: change scr... |
934 935 936 937 |
goto out; } BUG_ON(failed_mirror_index >= BTRFS_MAX_MIRRORS); sblock_bad = sblocks_for_recheck + failed_mirror_index; |
13db62b7a btrfs scrub: adde... |
938 |
|
b5d67f64f Btrfs: change scr... |
939 |
/* build and submit the bios for the failed mirror, check checksums */ |
34f5c8e90 Btrfs: in scrub r... |
940 |
scrub_recheck_block(fs_info, sblock_bad, is_metadata, have_csum, |
af8e2d1df Btrfs, scrub: rep... |
941 |
csum, generation, sctx->csum_size, 1); |
a2de733c7 btrfs: scrub |
942 |
|
b5d67f64f Btrfs: change scr... |
943 944 945 946 947 948 949 950 951 952 |
if (!sblock_bad->header_error && !sblock_bad->checksum_error && sblock_bad->no_io_error_seen) { /* * the error disappeared after reading page by page, or * the area was part of a huge bio and other parts of the * bio caused I/O errors, or the block layer merged several * read requests into one and the error is caused by a * different bio (usually one of the two latter cases is * the cause) */ |
d9d181c1b Btrfs: rename the... |
953 954 |
spin_lock(&sctx->stat_lock); sctx->stat.unverified_errors++; |
5a6ac9eac Btrfs, raid56: su... |
955 |
sblock_to_check->data_corrected = 1; |
d9d181c1b Btrfs: rename the... |
956 |
spin_unlock(&sctx->stat_lock); |
a2de733c7 btrfs: scrub |
957 |
|
ff023aac3 Btrfs: add code t... |
958 959 |
if (sctx->is_dev_replace) scrub_write_block_to_dev_replace(sblock_bad); |
b5d67f64f Btrfs: change scr... |
960 |
goto out; |
a2de733c7 btrfs: scrub |
961 |
} |
a2de733c7 btrfs: scrub |
962 |
|
b5d67f64f Btrfs: change scr... |
963 |
if (!sblock_bad->no_io_error_seen) { |
d9d181c1b Btrfs: rename the... |
964 965 966 |
spin_lock(&sctx->stat_lock); sctx->stat.read_errors++; spin_unlock(&sctx->stat_lock); |
b5d67f64f Btrfs: change scr... |
967 968 |
if (__ratelimit(&_rs)) scrub_print_warning("i/o error", sblock_to_check); |
a36cf8b89 Btrfs: remove the... |
969 |
btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_READ_ERRS); |
b5d67f64f Btrfs: change scr... |
970 |
} else if (sblock_bad->checksum_error) { |
d9d181c1b Btrfs: rename the... |
971 972 973 |
spin_lock(&sctx->stat_lock); sctx->stat.csum_errors++; spin_unlock(&sctx->stat_lock); |
b5d67f64f Btrfs: change scr... |
974 975 |
if (__ratelimit(&_rs)) scrub_print_warning("checksum error", sblock_to_check); |
a36cf8b89 Btrfs: remove the... |
976 |
btrfs_dev_stat_inc_and_print(dev, |
442a4f630 Btrfs: add device... |
977 |
BTRFS_DEV_STAT_CORRUPTION_ERRS); |
b5d67f64f Btrfs: change scr... |
978 |
} else if (sblock_bad->header_error) { |
d9d181c1b Btrfs: rename the... |
979 980 981 |
spin_lock(&sctx->stat_lock); sctx->stat.verify_errors++; spin_unlock(&sctx->stat_lock); |
b5d67f64f Btrfs: change scr... |
982 983 984 |
if (__ratelimit(&_rs)) scrub_print_warning("checksum/header error", sblock_to_check); |
442a4f630 Btrfs: add device... |
985 |
if (sblock_bad->generation_error) |
a36cf8b89 Btrfs: remove the... |
986 |
btrfs_dev_stat_inc_and_print(dev, |
442a4f630 Btrfs: add device... |
987 988 |
BTRFS_DEV_STAT_GENERATION_ERRS); else |
a36cf8b89 Btrfs: remove the... |
989 |
btrfs_dev_stat_inc_and_print(dev, |
442a4f630 Btrfs: add device... |
990 |
BTRFS_DEV_STAT_CORRUPTION_ERRS); |
b5d67f64f Btrfs: change scr... |
991 |
} |
a2de733c7 btrfs: scrub |
992 |
|
33ef30add Btrfs: do not inc... |
993 994 995 996 |
if (sctx->readonly) { ASSERT(!sctx->is_dev_replace); goto out; } |
a2de733c7 btrfs: scrub |
997 |
|
b5d67f64f Btrfs: change scr... |
998 999 |
if (!is_metadata && !have_csum) { struct scrub_fixup_nodatasum *fixup_nodatasum; |
a2de733c7 btrfs: scrub |
1000 |
|
ff023aac3 Btrfs: add code t... |
1001 1002 |
nodatasum_case: WARN_ON(sctx->is_dev_replace); |
b5d67f64f Btrfs: change scr... |
1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 |
/* * !is_metadata and !have_csum, this means that the data * might not be COW'ed, that it might be modified * concurrently. The general strategy to work on the * commit root does not help in the case when COW is not * used. */ fixup_nodatasum = kzalloc(sizeof(*fixup_nodatasum), GFP_NOFS); if (!fixup_nodatasum) goto did_not_correct_error; |
d9d181c1b Btrfs: rename the... |
1013 |
fixup_nodatasum->sctx = sctx; |
a36cf8b89 Btrfs: remove the... |
1014 |
fixup_nodatasum->dev = dev; |
b5d67f64f Btrfs: change scr... |
1015 1016 1017 |
fixup_nodatasum->logical = logical; fixup_nodatasum->root = fs_info->extent_root; fixup_nodatasum->mirror_num = failed_mirror_index + 1; |
b6bfebc13 Btrfs: cleanup sc... |
1018 |
scrub_pending_trans_workers_inc(sctx); |
9e0af2376 Btrfs: fix task h... |
1019 1020 |
btrfs_init_work(&fixup_nodatasum->work, btrfs_scrub_helper, scrub_fixup_nodatasum, NULL, NULL); |
0339ef2f4 btrfs: Replace fs... |
1021 1022 |
btrfs_queue_work(fs_info->scrub_workers, &fixup_nodatasum->work); |
b5d67f64f Btrfs: change scr... |
1023 |
goto out; |
a2de733c7 btrfs: scrub |
1024 |
} |
b5d67f64f Btrfs: change scr... |
1025 1026 |
/* * now build and submit the bios for the other mirrors, check |
cb2ced73d Btrfs: in scrub r... |
1027 1028 |
* checksums. * First try to pick the mirror which is completely without I/O |
b5d67f64f Btrfs: change scr... |
1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 |
* errors and also does not have a checksum error. * If one is found, and if a checksum is present, the full block * that is known to contain an error is rewritten. Afterwards * the block is known to be corrected. * If a mirror is found which is completely correct, and no * checksum is present, only those pages are rewritten that had * an I/O error in the block to be repaired, since it cannot be * determined, which copy of the other pages is better (and it * could happen otherwise that a correct page would be * overwritten by a bad one). */ for (mirror_index = 0; mirror_index < BTRFS_MAX_MIRRORS && sblocks_for_recheck[mirror_index].page_count > 0; mirror_index++) { |
cb2ced73d Btrfs: in scrub r... |
1044 |
struct scrub_block *sblock_other; |
b5d67f64f Btrfs: change scr... |
1045 |
|
cb2ced73d Btrfs: in scrub r... |
1046 1047 1048 1049 1050 |
if (mirror_index == failed_mirror_index) continue; sblock_other = sblocks_for_recheck + mirror_index; /* build and submit the bios, check checksums */ |
34f5c8e90 Btrfs: in scrub r... |
1051 1052 |
scrub_recheck_block(fs_info, sblock_other, is_metadata, have_csum, csum, generation, |
af8e2d1df Btrfs, scrub: rep... |
1053 |
sctx->csum_size, 0); |
34f5c8e90 Btrfs: in scrub r... |
1054 1055 |
if (!sblock_other->header_error && |
b5d67f64f Btrfs: change scr... |
1056 1057 |
!sblock_other->checksum_error && sblock_other->no_io_error_seen) { |
ff023aac3 Btrfs: add code t... |
1058 1059 1060 1061 1062 1063 1064 1065 1066 |
if (sctx->is_dev_replace) { scrub_write_block_to_dev_replace(sblock_other); } else { int force_write = is_metadata || have_csum; ret = scrub_repair_block_from_good_copy( sblock_bad, sblock_other, force_write); } |
b5d67f64f Btrfs: change scr... |
1067 1068 1069 1070 |
if (0 == ret) goto corrected_error; } } |
a2de733c7 btrfs: scrub |
1071 1072 |
/* |
ff023aac3 Btrfs: add code t... |
1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 |
* for dev_replace, pick good pages and write to the target device. */ if (sctx->is_dev_replace) { success = 1; for (page_num = 0; page_num < sblock_bad->page_count; page_num++) { int sub_success; sub_success = 0; for (mirror_index = 0; mirror_index < BTRFS_MAX_MIRRORS && sblocks_for_recheck[mirror_index].page_count > 0; mirror_index++) { struct scrub_block *sblock_other = sblocks_for_recheck + mirror_index; struct scrub_page *page_other = sblock_other->pagev[page_num]; if (!page_other->io_error) { ret = scrub_write_page_to_dev_replace( sblock_other, page_num); if (ret == 0) { /* succeeded for this page */ sub_success = 1; break; } else { btrfs_dev_replace_stats_inc( &sctx->dev_root-> fs_info->dev_replace. num_write_errors); } } } if (!sub_success) { /* * did not find a mirror to fetch the page * from. scrub_write_page_to_dev_replace() * handles this case (page->io_error), by * filling the block with zeros before * submitting the write request */ success = 0; ret = scrub_write_page_to_dev_replace( sblock_bad, page_num); if (ret) btrfs_dev_replace_stats_inc( &sctx->dev_root->fs_info-> dev_replace.num_write_errors); } } goto out; } /* * for regular scrub, repair those pages that are errored. * In case of I/O errors in the area that is supposed to be |
b5d67f64f Btrfs: change scr... |
1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 |
* repaired, continue by picking good copies of those pages. * Select the good pages from mirrors to rewrite bad pages from * the area to fix. Afterwards verify the checksum of the block * that is supposed to be repaired. This verification step is * only done for the purpose of statistic counting and for the * final scrub report, whether errors remain. * A perfect algorithm could make use of the checksum and try * all possible combinations of pages from the different mirrors * until the checksum verification succeeds. For example, when * the 2nd page of mirror #1 faces I/O errors, and the 2nd page * of mirror #2 is readable but the final checksum test fails, * then the 2nd page of mirror #3 could be tried, whether now * the final checksum succeedes. But this would be a rare * exception and is therefore not implemented. At least it is * avoided that the good copy is overwritten. * A more useful improvement would be to pick the sectors * without I/O error based on sector sizes (512 bytes on legacy * disks) instead of on PAGE_SIZE. Then maybe 512 byte of one * mirror could be repaired by taking 512 byte of a different * mirror, even if other 512 byte sectors in the same PAGE_SIZE * area are unreadable. |
a2de733c7 btrfs: scrub |
1152 |
*/ |
a2de733c7 btrfs: scrub |
1153 |
|
b5d67f64f Btrfs: change scr... |
1154 1155 1156 1157 1158 1159 |
/* can only fix I/O errors from here on */ if (sblock_bad->no_io_error_seen) goto did_not_correct_error; success = 1; for (page_num = 0; page_num < sblock_bad->page_count; page_num++) { |
7a9e99876 Btrfs: make the s... |
1160 |
struct scrub_page *page_bad = sblock_bad->pagev[page_num]; |
b5d67f64f Btrfs: change scr... |
1161 1162 |
if (!page_bad->io_error) |
a2de733c7 btrfs: scrub |
1163 |
continue; |
b5d67f64f Btrfs: change scr... |
1164 1165 1166 1167 1168 1169 1170 |
for (mirror_index = 0; mirror_index < BTRFS_MAX_MIRRORS && sblocks_for_recheck[mirror_index].page_count > 0; mirror_index++) { struct scrub_block *sblock_other = sblocks_for_recheck + mirror_index; |
7a9e99876 Btrfs: make the s... |
1171 1172 |
struct scrub_page *page_other = sblock_other->pagev[ page_num]; |
b5d67f64f Btrfs: change scr... |
1173 1174 1175 1176 1177 1178 1179 1180 1181 |
if (!page_other->io_error) { ret = scrub_repair_page_from_good_copy( sblock_bad, sblock_other, page_num, 0); if (0 == ret) { page_bad->io_error = 0; break; /* succeeded for this page */ } } |
96e369208 btrfs scrub: make... |
1182 |
} |
a2de733c7 btrfs: scrub |
1183 |
|
b5d67f64f Btrfs: change scr... |
1184 1185 1186 1187 |
if (page_bad->io_error) { /* did not find a mirror to copy the page from */ success = 0; } |
a2de733c7 btrfs: scrub |
1188 |
} |
a2de733c7 btrfs: scrub |
1189 |
|
b5d67f64f Btrfs: change scr... |
1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 |
if (success) { if (is_metadata || have_csum) { /* * need to verify the checksum now that all * sectors on disk are repaired (the write * request for data to be repaired is on its way). * Just be lazy and use scrub_recheck_block() * which re-reads the data before the checksum * is verified, but most likely the data comes out * of the page cache. */ |
34f5c8e90 Btrfs: in scrub r... |
1201 1202 |
scrub_recheck_block(fs_info, sblock_bad, is_metadata, have_csum, csum, |
af8e2d1df Btrfs, scrub: rep... |
1203 |
generation, sctx->csum_size, 1); |
34f5c8e90 Btrfs: in scrub r... |
1204 |
if (!sblock_bad->header_error && |
b5d67f64f Btrfs: change scr... |
1205 1206 1207 1208 1209 1210 1211 |
!sblock_bad->checksum_error && sblock_bad->no_io_error_seen) goto corrected_error; else goto did_not_correct_error; } else { corrected_error: |
d9d181c1b Btrfs: rename the... |
1212 1213 |
spin_lock(&sctx->stat_lock); sctx->stat.corrected_errors++; |
5a6ac9eac Btrfs, raid56: su... |
1214 |
sblock_to_check->data_corrected = 1; |
d9d181c1b Btrfs: rename the... |
1215 |
spin_unlock(&sctx->stat_lock); |
606686eea Btrfs: use rcu to... |
1216 |
printk_ratelimited_in_rcu(KERN_ERR |
efe120a06 Btrfs: convert pr... |
1217 1218 |
"BTRFS: fixed up error at logical %llu on dev %s ", |
c1c9ff7c9 Btrfs: Remove sup... |
1219 |
logical, rcu_str_deref(dev->name)); |
8628764e1 btrfs: add readon... |
1220 |
} |
b5d67f64f Btrfs: change scr... |
1221 1222 |
} else { did_not_correct_error: |
d9d181c1b Btrfs: rename the... |
1223 1224 1225 |
spin_lock(&sctx->stat_lock); sctx->stat.uncorrectable_errors++; spin_unlock(&sctx->stat_lock); |
606686eea Btrfs: use rcu to... |
1226 |
printk_ratelimited_in_rcu(KERN_ERR |
efe120a06 Btrfs: convert pr... |
1227 1228 |
"BTRFS: unable to fixup (regular) error at logical %llu on dev %s ", |
c1c9ff7c9 Btrfs: Remove sup... |
1229 |
logical, rcu_str_deref(dev->name)); |
96e369208 btrfs scrub: make... |
1230 |
} |
a2de733c7 btrfs: scrub |
1231 |
|
b5d67f64f Btrfs: change scr... |
1232 1233 1234 1235 1236 1237 |
out: if (sblocks_for_recheck) { for (mirror_index = 0; mirror_index < BTRFS_MAX_MIRRORS; mirror_index++) { struct scrub_block *sblock = sblocks_for_recheck + mirror_index; |
af8e2d1df Btrfs, scrub: rep... |
1238 |
struct scrub_recover *recover; |
b5d67f64f Btrfs: change scr... |
1239 |
int page_index; |
7a9e99876 Btrfs: make the s... |
1240 1241 1242 |
for (page_index = 0; page_index < sblock->page_count; page_index++) { sblock->pagev[page_index]->sblock = NULL; |
af8e2d1df Btrfs, scrub: rep... |
1243 1244 1245 1246 1247 1248 |
recover = sblock->pagev[page_index]->recover; if (recover) { scrub_put_recover(recover); sblock->pagev[page_index]->recover = NULL; } |
7a9e99876 Btrfs: make the s... |
1249 1250 |
scrub_page_put(sblock->pagev[page_index]); } |
b5d67f64f Btrfs: change scr... |
1251 1252 1253 |
} kfree(sblocks_for_recheck); } |
a2de733c7 btrfs: scrub |
1254 |
|
b5d67f64f Btrfs: change scr... |
1255 1256 |
return 0; } |
a2de733c7 btrfs: scrub |
1257 |
|
af8e2d1df Btrfs, scrub: rep... |
1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 |
static inline int scrub_nr_raid_mirrors(struct btrfs_bio *bbio, u64 *raid_map) { if (raid_map) { if (raid_map[bbio->num_stripes - 1] == RAID6_Q_STRIPE) return 3; else return 2; } else { return (int)bbio->num_stripes; } } static inline void scrub_stripe_index_and_offset(u64 logical, u64 *raid_map, u64 mapped_length, int nstripes, int mirror, int *stripe_index, u64 *stripe_offset) { int i; if (raid_map) { /* RAID5/6 */ for (i = 0; i < nstripes; i++) { if (raid_map[i] == RAID6_Q_STRIPE || raid_map[i] == RAID5_P_STRIPE) continue; if (logical >= raid_map[i] && logical < raid_map[i] + mapped_length) break; } *stripe_index = i; *stripe_offset = logical - raid_map[i]; } else { /* The other RAID type */ *stripe_index = mirror; *stripe_offset = 0; } } |
d9d181c1b Btrfs: rename the... |
1298 |
static int scrub_setup_recheck_block(struct scrub_ctx *sctx, |
3ec706c83 Btrfs: pass fs_in... |
1299 |
struct btrfs_fs_info *fs_info, |
ff023aac3 Btrfs: add code t... |
1300 |
struct scrub_block *original_sblock, |
b5d67f64f Btrfs: change scr... |
1301 1302 1303 |
u64 length, u64 logical, struct scrub_block *sblocks_for_recheck) { |
af8e2d1df Btrfs, scrub: rep... |
1304 1305 1306 1307 1308 1309 1310 |
struct scrub_recover *recover; struct btrfs_bio *bbio; u64 *raid_map; u64 sublen; u64 mapped_length; u64 stripe_offset; int stripe_index; |
b5d67f64f Btrfs: change scr... |
1311 1312 |
int page_index; int mirror_index; |
af8e2d1df Btrfs, scrub: rep... |
1313 |
int nmirrors; |
b5d67f64f Btrfs: change scr... |
1314 1315 1316 |
int ret; /* |
7a9e99876 Btrfs: make the s... |
1317 |
* note: the two members ref_count and outstanding_pages |
b5d67f64f Btrfs: change scr... |
1318 1319 1320 1321 1322 1323 |
* are not used (and not set) in the blocks that are used for * the recheck procedure */ page_index = 0; while (length > 0) { |
af8e2d1df Btrfs, scrub: rep... |
1324 1325 1326 1327 |
sublen = min_t(u64, length, PAGE_SIZE); mapped_length = sublen; bbio = NULL; raid_map = NULL; |
a2de733c7 btrfs: scrub |
1328 |
|
b5d67f64f Btrfs: change scr... |
1329 1330 1331 1332 |
/* * with a length of PAGE_SIZE, each returned stripe * represents one mirror */ |
af8e2d1df Btrfs, scrub: rep... |
1333 1334 |
ret = btrfs_map_sblock(fs_info, REQ_GET_READ_MIRRORS, logical, &mapped_length, &bbio, 0, &raid_map); |
b5d67f64f Btrfs: change scr... |
1335 1336 |
if (ret || !bbio || mapped_length < sublen) { kfree(bbio); |
af8e2d1df Btrfs, scrub: rep... |
1337 |
kfree(raid_map); |
b5d67f64f Btrfs: change scr... |
1338 1339 |
return -EIO; } |
a2de733c7 btrfs: scrub |
1340 |
|
af8e2d1df Btrfs, scrub: rep... |
1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 |
recover = kzalloc(sizeof(struct scrub_recover), GFP_NOFS); if (!recover) { kfree(bbio); kfree(raid_map); return -ENOMEM; } atomic_set(&recover->refs, 1); recover->bbio = bbio; recover->raid_map = raid_map; recover->map_length = mapped_length; |
ff023aac3 Btrfs: add code t... |
1352 |
BUG_ON(page_index >= SCRUB_PAGES_PER_RD_BIO); |
af8e2d1df Btrfs, scrub: rep... |
1353 1354 1355 |
nmirrors = scrub_nr_raid_mirrors(bbio, raid_map); for (mirror_index = 0; mirror_index < nmirrors; |
b5d67f64f Btrfs: change scr... |
1356 1357 1358 1359 1360 1361 1362 1363 |
mirror_index++) { struct scrub_block *sblock; struct scrub_page *page; if (mirror_index >= BTRFS_MAX_MIRRORS) continue; sblock = sblocks_for_recheck + mirror_index; |
7a9e99876 Btrfs: make the s... |
1364 1365 1366 1367 |
sblock->sctx = sctx; page = kzalloc(sizeof(*page), GFP_NOFS); if (!page) { leave_nomem: |
d9d181c1b Btrfs: rename the... |
1368 1369 1370 |
spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); |
af8e2d1df Btrfs, scrub: rep... |
1371 |
scrub_put_recover(recover); |
b5d67f64f Btrfs: change scr... |
1372 1373 |
return -ENOMEM; } |
7a9e99876 Btrfs: make the s... |
1374 1375 1376 |
scrub_page_get(page); sblock->pagev[page_index] = page; page->logical = logical; |
af8e2d1df Btrfs, scrub: rep... |
1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 |
scrub_stripe_index_and_offset(logical, raid_map, mapped_length, bbio->num_stripes, mirror_index, &stripe_index, &stripe_offset); page->physical = bbio->stripes[stripe_index].physical + stripe_offset; page->dev = bbio->stripes[stripe_index].dev; |
ff023aac3 Btrfs: add code t... |
1387 1388 1389 1390 |
BUG_ON(page_index >= original_sblock->page_count); page->physical_for_dev_replace = original_sblock->pagev[page_index]-> physical_for_dev_replace; |
7a9e99876 Btrfs: make the s... |
1391 |
/* for missing devices, dev->bdev is NULL */ |
7a9e99876 Btrfs: make the s... |
1392 |
page->mirror_num = mirror_index + 1; |
b5d67f64f Btrfs: change scr... |
1393 |
sblock->page_count++; |
7a9e99876 Btrfs: make the s... |
1394 1395 1396 |
page->page = alloc_page(GFP_NOFS); if (!page->page) goto leave_nomem; |
af8e2d1df Btrfs, scrub: rep... |
1397 1398 1399 |
scrub_get_recover(recover); page->recover = recover; |
b5d67f64f Btrfs: change scr... |
1400 |
} |
af8e2d1df Btrfs, scrub: rep... |
1401 |
scrub_put_recover(recover); |
b5d67f64f Btrfs: change scr... |
1402 1403 1404 1405 1406 1407 |
length -= sublen; logical += sublen; page_index++; } return 0; |
96e369208 btrfs scrub: make... |
1408 |
} |
af8e2d1df Btrfs, scrub: rep... |
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 |
struct scrub_bio_ret { struct completion event; int error; }; static void scrub_bio_wait_endio(struct bio *bio, int error) { struct scrub_bio_ret *ret = bio->bi_private; ret->error = error; complete(&ret->event); } static inline int scrub_is_page_on_raid56(struct scrub_page *page) { return page->recover && page->recover->raid_map; } static int scrub_submit_raid56_bio_wait(struct btrfs_fs_info *fs_info, struct bio *bio, struct scrub_page *page) { struct scrub_bio_ret done; int ret; init_completion(&done.event); done.error = 0; bio->bi_iter.bi_sector = page->logical >> 9; bio->bi_private = &done; bio->bi_end_io = scrub_bio_wait_endio; ret = raid56_parity_recover(fs_info->fs_root, bio, page->recover->bbio, page->recover->raid_map, page->recover->map_length, |
4245215d6 Btrfs, raid56: fi... |
1443 |
page->mirror_num, 0); |
af8e2d1df Btrfs, scrub: rep... |
1444 1445 1446 1447 1448 1449 1450 1451 1452 |
if (ret) return ret; wait_for_completion(&done.event); if (done.error) return -EIO; return 0; } |
b5d67f64f Btrfs: change scr... |
1453 1454 1455 1456 1457 1458 1459 |
/* * this function will check the on disk data for checksum errors, header * errors and read I/O errors. If any I/O errors happen, the exact pages * which are errored are marked as being bad. The goal is to enable scrub * to take those pages that are not errored from all the mirrors so that * the pages that are errored in the just handled mirror can be repaired. */ |
34f5c8e90 Btrfs: in scrub r... |
1460 1461 1462 |
static void scrub_recheck_block(struct btrfs_fs_info *fs_info, struct scrub_block *sblock, int is_metadata, int have_csum, u8 *csum, u64 generation, |
af8e2d1df Btrfs, scrub: rep... |
1463 |
u16 csum_size, int retry_failed_mirror) |
96e369208 btrfs scrub: make... |
1464 |
{ |
b5d67f64f Btrfs: change scr... |
1465 |
int page_num; |
96e369208 btrfs scrub: make... |
1466 |
|
b5d67f64f Btrfs: change scr... |
1467 1468 1469 |
sblock->no_io_error_seen = 1; sblock->header_error = 0; sblock->checksum_error = 0; |
96e369208 btrfs scrub: make... |
1470 |
|
b5d67f64f Btrfs: change scr... |
1471 1472 |
for (page_num = 0; page_num < sblock->page_count; page_num++) { struct bio *bio; |
7a9e99876 Btrfs: make the s... |
1473 |
struct scrub_page *page = sblock->pagev[page_num]; |
b5d67f64f Btrfs: change scr... |
1474 |
|
442a4f630 Btrfs: add device... |
1475 |
if (page->dev->bdev == NULL) { |
ea9947b43 Btrfs: fix crash ... |
1476 1477 1478 1479 |
page->io_error = 1; sblock->no_io_error_seen = 0; continue; } |
7a9e99876 Btrfs: make the s... |
1480 |
WARN_ON(!page->page); |
9be3395bc Btrfs: use a btrf... |
1481 |
bio = btrfs_io_bio_alloc(GFP_NOFS, 1); |
34f5c8e90 Btrfs: in scrub r... |
1482 1483 1484 1485 1486 |
if (!bio) { page->io_error = 1; sblock->no_io_error_seen = 0; continue; } |
442a4f630 Btrfs: add device... |
1487 |
bio->bi_bdev = page->dev->bdev; |
b5d67f64f Btrfs: change scr... |
1488 |
|
34f5c8e90 Btrfs: in scrub r... |
1489 |
bio_add_page(bio, page->page, PAGE_SIZE, 0); |
af8e2d1df Btrfs, scrub: rep... |
1490 1491 1492 1493 1494 1495 1496 1497 1498 |
if (!retry_failed_mirror && scrub_is_page_on_raid56(page)) { if (scrub_submit_raid56_bio_wait(fs_info, bio, page)) sblock->no_io_error_seen = 0; } else { bio->bi_iter.bi_sector = page->physical >> 9; if (btrfsic_submit_bio_wait(READ, bio)) sblock->no_io_error_seen = 0; } |
33879d451 block: submit_bio... |
1499 |
|
b5d67f64f Btrfs: change scr... |
1500 1501 |
bio_put(bio); } |
96e369208 btrfs scrub: make... |
1502 |
|
b5d67f64f Btrfs: change scr... |
1503 1504 1505 1506 |
if (sblock->no_io_error_seen) scrub_recheck_block_checksum(fs_info, sblock, is_metadata, have_csum, csum, generation, csum_size); |
34f5c8e90 Btrfs: in scrub r... |
1507 |
return; |
a2de733c7 btrfs: scrub |
1508 |
} |
17a9be2f2 Btrfs: fix wrong ... |
1509 1510 1511 1512 1513 1514 1515 1516 1517 |
static inline int scrub_check_fsid(u8 fsid[], struct scrub_page *spage) { struct btrfs_fs_devices *fs_devices = spage->dev->fs_devices; int ret; ret = memcmp(fsid, fs_devices->fsid, BTRFS_UUID_SIZE); return !ret; } |
b5d67f64f Btrfs: change scr... |
1518 1519 1520 1521 1522 |
static void scrub_recheck_block_checksum(struct btrfs_fs_info *fs_info, struct scrub_block *sblock, int is_metadata, int have_csum, const u8 *csum, u64 generation, u16 csum_size) |
a2de733c7 btrfs: scrub |
1523 |
{ |
b5d67f64f Btrfs: change scr... |
1524 1525 1526 |
int page_num; u8 calculated_csum[BTRFS_CSUM_SIZE]; u32 crc = ~(u32)0; |
b5d67f64f Btrfs: change scr... |
1527 |
void *mapped_buffer; |
7a9e99876 Btrfs: make the s... |
1528 |
WARN_ON(!sblock->pagev[0]->page); |
b5d67f64f Btrfs: change scr... |
1529 1530 |
if (is_metadata) { struct btrfs_header *h; |
7a9e99876 Btrfs: make the s... |
1531 |
mapped_buffer = kmap_atomic(sblock->pagev[0]->page); |
b5d67f64f Btrfs: change scr... |
1532 |
h = (struct btrfs_header *)mapped_buffer; |
3cae210fa btrfs: Cleanup fo... |
1533 |
if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h) || |
17a9be2f2 Btrfs: fix wrong ... |
1534 |
!scrub_check_fsid(h->fsid, sblock->pagev[0]) || |
b5d67f64f Btrfs: change scr... |
1535 |
memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid, |
442a4f630 Btrfs: add device... |
1536 |
BTRFS_UUID_SIZE)) { |
b5d67f64f Btrfs: change scr... |
1537 |
sblock->header_error = 1; |
3cae210fa btrfs: Cleanup fo... |
1538 |
} else if (generation != btrfs_stack_header_generation(h)) { |
442a4f630 Btrfs: add device... |
1539 1540 1541 |
sblock->header_error = 1; sblock->generation_error = 1; } |
b5d67f64f Btrfs: change scr... |
1542 1543 1544 1545 |
csum = h->csum; } else { if (!have_csum) return; |
a2de733c7 btrfs: scrub |
1546 |
|
7a9e99876 Btrfs: make the s... |
1547 |
mapped_buffer = kmap_atomic(sblock->pagev[0]->page); |
b5d67f64f Btrfs: change scr... |
1548 |
} |
a2de733c7 btrfs: scrub |
1549 |
|
b5d67f64f Btrfs: change scr... |
1550 1551 |
for (page_num = 0;;) { if (page_num == 0 && is_metadata) |
b0496686b Btrfs: cleanup un... |
1552 |
crc = btrfs_csum_data( |
b5d67f64f Btrfs: change scr... |
1553 1554 1555 |
((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE, crc, PAGE_SIZE - BTRFS_CSUM_SIZE); else |
b0496686b Btrfs: cleanup un... |
1556 |
crc = btrfs_csum_data(mapped_buffer, crc, PAGE_SIZE); |
b5d67f64f Btrfs: change scr... |
1557 |
|
9613bebb2 Merge branch 'for... |
1558 |
kunmap_atomic(mapped_buffer); |
b5d67f64f Btrfs: change scr... |
1559 1560 1561 |
page_num++; if (page_num >= sblock->page_count) break; |
7a9e99876 Btrfs: make the s... |
1562 |
WARN_ON(!sblock->pagev[page_num]->page); |
b5d67f64f Btrfs: change scr... |
1563 |
|
7a9e99876 Btrfs: make the s... |
1564 |
mapped_buffer = kmap_atomic(sblock->pagev[page_num]->page); |
b5d67f64f Btrfs: change scr... |
1565 1566 1567 1568 1569 |
} btrfs_csum_final(crc, calculated_csum); if (memcmp(calculated_csum, csum, csum_size)) sblock->checksum_error = 1; |
a2de733c7 btrfs: scrub |
1570 |
} |
b5d67f64f Btrfs: change scr... |
1571 1572 1573 1574 1575 1576 |
static int scrub_repair_block_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int force_write) { int page_num; int ret = 0; |
96e369208 btrfs scrub: make... |
1577 |
|
b5d67f64f Btrfs: change scr... |
1578 1579 |
for (page_num = 0; page_num < sblock_bad->page_count; page_num++) { int ret_sub; |
96e369208 btrfs scrub: make... |
1580 |
|
b5d67f64f Btrfs: change scr... |
1581 1582 1583 1584 1585 1586 |
ret_sub = scrub_repair_page_from_good_copy(sblock_bad, sblock_good, page_num, force_write); if (ret_sub) ret = ret_sub; |
a2de733c7 btrfs: scrub |
1587 |
} |
b5d67f64f Btrfs: change scr... |
1588 1589 1590 1591 1592 1593 1594 1595 |
return ret; } static int scrub_repair_page_from_good_copy(struct scrub_block *sblock_bad, struct scrub_block *sblock_good, int page_num, int force_write) { |
7a9e99876 Btrfs: make the s... |
1596 1597 |
struct scrub_page *page_bad = sblock_bad->pagev[page_num]; struct scrub_page *page_good = sblock_good->pagev[page_num]; |
b5d67f64f Btrfs: change scr... |
1598 |
|
7a9e99876 Btrfs: make the s... |
1599 1600 |
BUG_ON(page_bad->page == NULL); BUG_ON(page_good->page == NULL); |
b5d67f64f Btrfs: change scr... |
1601 1602 1603 1604 |
if (force_write || sblock_bad->header_error || sblock_bad->checksum_error || page_bad->io_error) { struct bio *bio; int ret; |
b5d67f64f Btrfs: change scr... |
1605 |
|
ff023aac3 Btrfs: add code t... |
1606 |
if (!page_bad->dev->bdev) { |
efe120a06 Btrfs: convert pr... |
1607 1608 1609 1610 |
printk_ratelimited(KERN_WARNING "BTRFS: " "scrub_repair_page_from_good_copy(bdev == NULL) " "is unexpected! "); |
ff023aac3 Btrfs: add code t... |
1611 1612 |
return -EIO; } |
9be3395bc Btrfs: use a btrf... |
1613 |
bio = btrfs_io_bio_alloc(GFP_NOFS, 1); |
e627ee7bc Btrfs: check retu... |
1614 1615 |
if (!bio) return -EIO; |
442a4f630 Btrfs: add device... |
1616 |
bio->bi_bdev = page_bad->dev->bdev; |
4f024f379 block: Abstract o... |
1617 |
bio->bi_iter.bi_sector = page_bad->physical >> 9; |
b5d67f64f Btrfs: change scr... |
1618 1619 1620 1621 1622 |
ret = bio_add_page(bio, page_good->page, PAGE_SIZE, 0); if (PAGE_SIZE != ret) { bio_put(bio); return -EIO; |
13db62b7a btrfs scrub: adde... |
1623 |
} |
b5d67f64f Btrfs: change scr... |
1624 |
|
33879d451 block: submit_bio... |
1625 |
if (btrfsic_submit_bio_wait(WRITE, bio)) { |
442a4f630 Btrfs: add device... |
1626 1627 |
btrfs_dev_stat_inc_and_print(page_bad->dev, BTRFS_DEV_STAT_WRITE_ERRS); |
ff023aac3 Btrfs: add code t... |
1628 1629 1630 |
btrfs_dev_replace_stats_inc( &sblock_bad->sctx->dev_root->fs_info-> dev_replace.num_write_errors); |
442a4f630 Btrfs: add device... |
1631 1632 1633 |
bio_put(bio); return -EIO; } |
b5d67f64f Btrfs: change scr... |
1634 |
bio_put(bio); |
a2de733c7 btrfs: scrub |
1635 |
} |
b5d67f64f Btrfs: change scr... |
1636 1637 |
return 0; } |
ff023aac3 Btrfs: add code t... |
1638 1639 1640 |
static void scrub_write_block_to_dev_replace(struct scrub_block *sblock) { int page_num; |
5a6ac9eac Btrfs, raid56: su... |
1641 1642 1643 1644 1645 1646 |
/* * This block is used for the check of the parity on the source device, * so the data needn't be written into the destination device. */ if (sblock->sparity) return; |
ff023aac3 Btrfs: add code t... |
1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 |
for (page_num = 0; page_num < sblock->page_count; page_num++) { int ret; ret = scrub_write_page_to_dev_replace(sblock, page_num); if (ret) btrfs_dev_replace_stats_inc( &sblock->sctx->dev_root->fs_info->dev_replace. num_write_errors); } } static int scrub_write_page_to_dev_replace(struct scrub_block *sblock, int page_num) { struct scrub_page *spage = sblock->pagev[page_num]; BUG_ON(spage->page == NULL); if (spage->io_error) { void *mapped_buffer = kmap_atomic(spage->page); memset(mapped_buffer, 0, PAGE_CACHE_SIZE); flush_dcache_page(spage->page); kunmap_atomic(mapped_buffer); } return scrub_add_page_to_wr_bio(sblock->sctx, spage); } static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx, struct scrub_page *spage) { struct scrub_wr_ctx *wr_ctx = &sctx->wr_ctx; struct scrub_bio *sbio; int ret; mutex_lock(&wr_ctx->wr_lock); again: if (!wr_ctx->wr_curr_bio) { wr_ctx->wr_curr_bio = kzalloc(sizeof(*wr_ctx->wr_curr_bio), GFP_NOFS); if (!wr_ctx->wr_curr_bio) { mutex_unlock(&wr_ctx->wr_lock); return -ENOMEM; } wr_ctx->wr_curr_bio->sctx = sctx; wr_ctx->wr_curr_bio->page_count = 0; } sbio = wr_ctx->wr_curr_bio; if (sbio->page_count == 0) { struct bio *bio; sbio->physical = spage->physical_for_dev_replace; sbio->logical = spage->logical; sbio->dev = wr_ctx->tgtdev; bio = sbio->bio; if (!bio) { |
9be3395bc Btrfs: use a btrf... |
1702 |
bio = btrfs_io_bio_alloc(GFP_NOFS, wr_ctx->pages_per_wr_bio); |
ff023aac3 Btrfs: add code t... |
1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 |
if (!bio) { mutex_unlock(&wr_ctx->wr_lock); return -ENOMEM; } sbio->bio = bio; } bio->bi_private = sbio; bio->bi_end_io = scrub_wr_bio_end_io; bio->bi_bdev = sbio->dev->bdev; |
4f024f379 block: Abstract o... |
1713 |
bio->bi_iter.bi_sector = sbio->physical >> 9; |
ff023aac3 Btrfs: add code t... |
1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 |
sbio->err = 0; } else if (sbio->physical + sbio->page_count * PAGE_SIZE != spage->physical_for_dev_replace || sbio->logical + sbio->page_count * PAGE_SIZE != spage->logical) { scrub_wr_submit(sctx); goto again; } ret = bio_add_page(sbio->bio, spage->page, PAGE_SIZE, 0); if (ret != PAGE_SIZE) { if (sbio->page_count < 1) { bio_put(sbio->bio); sbio->bio = NULL; mutex_unlock(&wr_ctx->wr_lock); return -EIO; } scrub_wr_submit(sctx); goto again; } sbio->pagev[sbio->page_count] = spage; scrub_page_get(spage); sbio->page_count++; if (sbio->page_count == wr_ctx->pages_per_wr_bio) scrub_wr_submit(sctx); mutex_unlock(&wr_ctx->wr_lock); return 0; } static void scrub_wr_submit(struct scrub_ctx *sctx) { struct scrub_wr_ctx *wr_ctx = &sctx->wr_ctx; struct scrub_bio *sbio; if (!wr_ctx->wr_curr_bio) return; sbio = wr_ctx->wr_curr_bio; wr_ctx->wr_curr_bio = NULL; WARN_ON(!sbio->bio->bi_bdev); scrub_pending_bio_inc(sctx); /* process all writes in a single worker thread. Then the block layer * orders the requests before sending them to the driver which * doubled the write performance on spinning disks when measured * with Linux 3.5 */ btrfsic_submit_bio(WRITE, sbio->bio); } static void scrub_wr_bio_end_io(struct bio *bio, int err) { struct scrub_bio *sbio = bio->bi_private; struct btrfs_fs_info *fs_info = sbio->dev->dev_root->fs_info; sbio->err = err; sbio->bio = bio; |
9e0af2376 Btrfs: fix task h... |
1771 1772 |
btrfs_init_work(&sbio->work, btrfs_scrubwrc_helper, scrub_wr_bio_end_io_worker, NULL, NULL); |
0339ef2f4 btrfs: Replace fs... |
1773 |
btrfs_queue_work(fs_info->scrub_wr_completion_workers, &sbio->work); |
ff023aac3 Btrfs: add code t... |
1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 |
} static void scrub_wr_bio_end_io_worker(struct btrfs_work *work) { struct scrub_bio *sbio = container_of(work, struct scrub_bio, work); struct scrub_ctx *sctx = sbio->sctx; int i; WARN_ON(sbio->page_count > SCRUB_PAGES_PER_WR_BIO); if (sbio->err) { struct btrfs_dev_replace *dev_replace = &sbio->sctx->dev_root->fs_info->dev_replace; for (i = 0; i < sbio->page_count; i++) { struct scrub_page *spage = sbio->pagev[i]; spage->io_error = 1; btrfs_dev_replace_stats_inc(&dev_replace-> num_write_errors); } } for (i = 0; i < sbio->page_count; i++) scrub_page_put(sbio->pagev[i]); bio_put(sbio->bio); kfree(sbio); scrub_pending_bio_dec(sctx); } static int scrub_checksum(struct scrub_block *sblock) |
b5d67f64f Btrfs: change scr... |
1805 1806 1807 |
{ u64 flags; int ret; |
7a9e99876 Btrfs: make the s... |
1808 1809 |
WARN_ON(sblock->page_count < 1); flags = sblock->pagev[0]->flags; |
b5d67f64f Btrfs: change scr... |
1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 |
ret = 0; if (flags & BTRFS_EXTENT_FLAG_DATA) ret = scrub_checksum_data(sblock); else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) ret = scrub_checksum_tree_block(sblock); else if (flags & BTRFS_EXTENT_FLAG_SUPER) (void)scrub_checksum_super(sblock); else WARN_ON(1); if (ret) scrub_handle_errored_block(sblock); |
ff023aac3 Btrfs: add code t... |
1821 1822 |
return ret; |
a2de733c7 btrfs: scrub |
1823 |
} |
b5d67f64f Btrfs: change scr... |
1824 |
static int scrub_checksum_data(struct scrub_block *sblock) |
a2de733c7 btrfs: scrub |
1825 |
{ |
d9d181c1b Btrfs: rename the... |
1826 |
struct scrub_ctx *sctx = sblock->sctx; |
a2de733c7 btrfs: scrub |
1827 |
u8 csum[BTRFS_CSUM_SIZE]; |
b5d67f64f Btrfs: change scr... |
1828 1829 1830 |
u8 *on_disk_csum; struct page *page; void *buffer; |
a2de733c7 btrfs: scrub |
1831 1832 |
u32 crc = ~(u32)0; int fail = 0; |
b5d67f64f Btrfs: change scr... |
1833 1834 |
u64 len; int index; |
a2de733c7 btrfs: scrub |
1835 |
|
b5d67f64f Btrfs: change scr... |
1836 |
BUG_ON(sblock->page_count < 1); |
7a9e99876 Btrfs: make the s... |
1837 |
if (!sblock->pagev[0]->have_csum) |
a2de733c7 btrfs: scrub |
1838 |
return 0; |
7a9e99876 Btrfs: make the s... |
1839 1840 |
on_disk_csum = sblock->pagev[0]->csum; page = sblock->pagev[0]->page; |
9613bebb2 Merge branch 'for... |
1841 |
buffer = kmap_atomic(page); |
b5d67f64f Btrfs: change scr... |
1842 |
|
d9d181c1b Btrfs: rename the... |
1843 |
len = sctx->sectorsize; |
b5d67f64f Btrfs: change scr... |
1844 1845 1846 |
index = 0; for (;;) { u64 l = min_t(u64, len, PAGE_SIZE); |
b0496686b Btrfs: cleanup un... |
1847 |
crc = btrfs_csum_data(buffer, crc, l); |
9613bebb2 Merge branch 'for... |
1848 |
kunmap_atomic(buffer); |
b5d67f64f Btrfs: change scr... |
1849 1850 1851 1852 1853 |
len -= l; if (len == 0) break; index++; BUG_ON(index >= sblock->page_count); |
7a9e99876 Btrfs: make the s... |
1854 1855 |
BUG_ON(!sblock->pagev[index]->page); page = sblock->pagev[index]->page; |
9613bebb2 Merge branch 'for... |
1856 |
buffer = kmap_atomic(page); |
b5d67f64f Btrfs: change scr... |
1857 |
} |
a2de733c7 btrfs: scrub |
1858 |
btrfs_csum_final(crc, csum); |
d9d181c1b Btrfs: rename the... |
1859 |
if (memcmp(csum, on_disk_csum, sctx->csum_size)) |
a2de733c7 btrfs: scrub |
1860 |
fail = 1; |
a2de733c7 btrfs: scrub |
1861 1862 |
return fail; } |
b5d67f64f Btrfs: change scr... |
1863 |
static int scrub_checksum_tree_block(struct scrub_block *sblock) |
a2de733c7 btrfs: scrub |
1864 |
{ |
d9d181c1b Btrfs: rename the... |
1865 |
struct scrub_ctx *sctx = sblock->sctx; |
a2de733c7 btrfs: scrub |
1866 |
struct btrfs_header *h; |
a36cf8b89 Btrfs: remove the... |
1867 |
struct btrfs_root *root = sctx->dev_root; |
a2de733c7 btrfs: scrub |
1868 |
struct btrfs_fs_info *fs_info = root->fs_info; |
b5d67f64f Btrfs: change scr... |
1869 1870 1871 1872 1873 1874 |
u8 calculated_csum[BTRFS_CSUM_SIZE]; u8 on_disk_csum[BTRFS_CSUM_SIZE]; struct page *page; void *mapped_buffer; u64 mapped_size; void *p; |
a2de733c7 btrfs: scrub |
1875 1876 1877 |
u32 crc = ~(u32)0; int fail = 0; int crc_fail = 0; |
b5d67f64f Btrfs: change scr... |
1878 1879 1880 1881 |
u64 len; int index; BUG_ON(sblock->page_count < 1); |
7a9e99876 Btrfs: make the s... |
1882 |
page = sblock->pagev[0]->page; |
9613bebb2 Merge branch 'for... |
1883 |
mapped_buffer = kmap_atomic(page); |
b5d67f64f Btrfs: change scr... |
1884 |
h = (struct btrfs_header *)mapped_buffer; |
d9d181c1b Btrfs: rename the... |
1885 |
memcpy(on_disk_csum, h->csum, sctx->csum_size); |
a2de733c7 btrfs: scrub |
1886 1887 1888 1889 1890 1891 |
/* * we don't use the getter functions here, as we * a) don't have an extent buffer and * b) the page is already kmapped */ |
a2de733c7 btrfs: scrub |
1892 |
|
3cae210fa btrfs: Cleanup fo... |
1893 |
if (sblock->pagev[0]->logical != btrfs_stack_header_bytenr(h)) |
a2de733c7 btrfs: scrub |
1894 |
++fail; |
3cae210fa btrfs: Cleanup fo... |
1895 |
if (sblock->pagev[0]->generation != btrfs_stack_header_generation(h)) |
a2de733c7 btrfs: scrub |
1896 |
++fail; |
17a9be2f2 Btrfs: fix wrong ... |
1897 |
if (!scrub_check_fsid(h->fsid, sblock->pagev[0])) |
a2de733c7 btrfs: scrub |
1898 1899 1900 1901 1902 |
++fail; if (memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid, BTRFS_UUID_SIZE)) ++fail; |
d9d181c1b Btrfs: rename the... |
1903 |
len = sctx->nodesize - BTRFS_CSUM_SIZE; |
b5d67f64f Btrfs: change scr... |
1904 1905 1906 1907 1908 |
mapped_size = PAGE_SIZE - BTRFS_CSUM_SIZE; p = ((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE; index = 0; for (;;) { u64 l = min_t(u64, len, mapped_size); |
b0496686b Btrfs: cleanup un... |
1909 |
crc = btrfs_csum_data(p, crc, l); |
9613bebb2 Merge branch 'for... |
1910 |
kunmap_atomic(mapped_buffer); |
b5d67f64f Btrfs: change scr... |
1911 1912 1913 1914 1915 |
len -= l; if (len == 0) break; index++; BUG_ON(index >= sblock->page_count); |
7a9e99876 Btrfs: make the s... |
1916 1917 |
BUG_ON(!sblock->pagev[index]->page); page = sblock->pagev[index]->page; |
9613bebb2 Merge branch 'for... |
1918 |
mapped_buffer = kmap_atomic(page); |
b5d67f64f Btrfs: change scr... |
1919 1920 1921 1922 1923 |
mapped_size = PAGE_SIZE; p = mapped_buffer; } btrfs_csum_final(crc, calculated_csum); |
d9d181c1b Btrfs: rename the... |
1924 |
if (memcmp(calculated_csum, on_disk_csum, sctx->csum_size)) |
a2de733c7 btrfs: scrub |
1925 |
++crc_fail; |
a2de733c7 btrfs: scrub |
1926 1927 |
return fail || crc_fail; } |
b5d67f64f Btrfs: change scr... |
1928 |
static int scrub_checksum_super(struct scrub_block *sblock) |
a2de733c7 btrfs: scrub |
1929 1930 |
{ struct btrfs_super_block *s; |
d9d181c1b Btrfs: rename the... |
1931 |
struct scrub_ctx *sctx = sblock->sctx; |
b5d67f64f Btrfs: change scr... |
1932 1933 1934 1935 1936 1937 |
u8 calculated_csum[BTRFS_CSUM_SIZE]; u8 on_disk_csum[BTRFS_CSUM_SIZE]; struct page *page; void *mapped_buffer; u64 mapped_size; void *p; |
a2de733c7 btrfs: scrub |
1938 |
u32 crc = ~(u32)0; |
442a4f630 Btrfs: add device... |
1939 1940 |
int fail_gen = 0; int fail_cor = 0; |
b5d67f64f Btrfs: change scr... |
1941 1942 |
u64 len; int index; |
a2de733c7 btrfs: scrub |
1943 |
|
b5d67f64f Btrfs: change scr... |
1944 |
BUG_ON(sblock->page_count < 1); |
7a9e99876 Btrfs: make the s... |
1945 |
page = sblock->pagev[0]->page; |
9613bebb2 Merge branch 'for... |
1946 |
mapped_buffer = kmap_atomic(page); |
b5d67f64f Btrfs: change scr... |
1947 |
s = (struct btrfs_super_block *)mapped_buffer; |
d9d181c1b Btrfs: rename the... |
1948 |
memcpy(on_disk_csum, s->csum, sctx->csum_size); |
a2de733c7 btrfs: scrub |
1949 |
|
3cae210fa btrfs: Cleanup fo... |
1950 |
if (sblock->pagev[0]->logical != btrfs_super_bytenr(s)) |
442a4f630 Btrfs: add device... |
1951 |
++fail_cor; |
a2de733c7 btrfs: scrub |
1952 |
|
3cae210fa btrfs: Cleanup fo... |
1953 |
if (sblock->pagev[0]->generation != btrfs_super_generation(s)) |
442a4f630 Btrfs: add device... |
1954 |
++fail_gen; |
a2de733c7 btrfs: scrub |
1955 |
|
17a9be2f2 Btrfs: fix wrong ... |
1956 |
if (!scrub_check_fsid(s->fsid, sblock->pagev[0])) |
442a4f630 Btrfs: add device... |
1957 |
++fail_cor; |
a2de733c7 btrfs: scrub |
1958 |
|
b5d67f64f Btrfs: change scr... |
1959 1960 1961 1962 1963 1964 |
len = BTRFS_SUPER_INFO_SIZE - BTRFS_CSUM_SIZE; mapped_size = PAGE_SIZE - BTRFS_CSUM_SIZE; p = ((u8 *)mapped_buffer) + BTRFS_CSUM_SIZE; index = 0; for (;;) { u64 l = min_t(u64, len, mapped_size); |
b0496686b Btrfs: cleanup un... |
1965 |
crc = btrfs_csum_data(p, crc, l); |
9613bebb2 Merge branch 'for... |
1966 |
kunmap_atomic(mapped_buffer); |
b5d67f64f Btrfs: change scr... |
1967 1968 1969 1970 1971 |
len -= l; if (len == 0) break; index++; BUG_ON(index >= sblock->page_count); |
7a9e99876 Btrfs: make the s... |
1972 1973 |
BUG_ON(!sblock->pagev[index]->page); page = sblock->pagev[index]->page; |
9613bebb2 Merge branch 'for... |
1974 |
mapped_buffer = kmap_atomic(page); |
b5d67f64f Btrfs: change scr... |
1975 1976 1977 1978 1979 |
mapped_size = PAGE_SIZE; p = mapped_buffer; } btrfs_csum_final(crc, calculated_csum); |
d9d181c1b Btrfs: rename the... |
1980 |
if (memcmp(calculated_csum, on_disk_csum, sctx->csum_size)) |
442a4f630 Btrfs: add device... |
1981 |
++fail_cor; |
a2de733c7 btrfs: scrub |
1982 |
|
442a4f630 Btrfs: add device... |
1983 |
if (fail_cor + fail_gen) { |
a2de733c7 btrfs: scrub |
1984 1985 1986 1987 1988 |
/* * if we find an error in a super block, we just report it. * They will get written with the next transaction commit * anyway */ |
d9d181c1b Btrfs: rename the... |
1989 1990 1991 |
spin_lock(&sctx->stat_lock); ++sctx->stat.super_errors; spin_unlock(&sctx->stat_lock); |
442a4f630 Btrfs: add device... |
1992 |
if (fail_cor) |
7a9e99876 Btrfs: make the s... |
1993 |
btrfs_dev_stat_inc_and_print(sblock->pagev[0]->dev, |
442a4f630 Btrfs: add device... |
1994 1995 |
BTRFS_DEV_STAT_CORRUPTION_ERRS); else |
7a9e99876 Btrfs: make the s... |
1996 |
btrfs_dev_stat_inc_and_print(sblock->pagev[0]->dev, |
442a4f630 Btrfs: add device... |
1997 |
BTRFS_DEV_STAT_GENERATION_ERRS); |
a2de733c7 btrfs: scrub |
1998 |
} |
442a4f630 Btrfs: add device... |
1999 |
return fail_cor + fail_gen; |
a2de733c7 btrfs: scrub |
2000 |
} |
b5d67f64f Btrfs: change scr... |
2001 2002 2003 2004 2005 2006 2007 2008 2009 |
static void scrub_block_get(struct scrub_block *sblock) { atomic_inc(&sblock->ref_count); } static void scrub_block_put(struct scrub_block *sblock) { if (atomic_dec_and_test(&sblock->ref_count)) { int i; |
5a6ac9eac Btrfs, raid56: su... |
2010 2011 |
if (sblock->sparity) scrub_parity_put(sblock->sparity); |
b5d67f64f Btrfs: change scr... |
2012 |
for (i = 0; i < sblock->page_count; i++) |
7a9e99876 Btrfs: make the s... |
2013 |
scrub_page_put(sblock->pagev[i]); |
b5d67f64f Btrfs: change scr... |
2014 2015 2016 |
kfree(sblock); } } |
7a9e99876 Btrfs: make the s... |
2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 |
static void scrub_page_get(struct scrub_page *spage) { atomic_inc(&spage->ref_count); } static void scrub_page_put(struct scrub_page *spage) { if (atomic_dec_and_test(&spage->ref_count)) { if (spage->page) __free_page(spage->page); kfree(spage); } } |
d9d181c1b Btrfs: rename the... |
2030 |
static void scrub_submit(struct scrub_ctx *sctx) |
a2de733c7 btrfs: scrub |
2031 2032 |
{ struct scrub_bio *sbio; |
d9d181c1b Btrfs: rename the... |
2033 |
if (sctx->curr == -1) |
1623edebe Btrfs: minor clea... |
2034 |
return; |
a2de733c7 btrfs: scrub |
2035 |
|
d9d181c1b Btrfs: rename the... |
2036 2037 |
sbio = sctx->bios[sctx->curr]; sctx->curr = -1; |
b6bfebc13 Btrfs: cleanup sc... |
2038 |
scrub_pending_bio_inc(sctx); |
a2de733c7 btrfs: scrub |
2039 |
|
ff023aac3 Btrfs: add code t... |
2040 2041 2042 2043 2044 2045 2046 2047 2048 |
if (!sbio->bio->bi_bdev) { /* * this case should not happen. If btrfs_map_block() is * wrong, it could happen for dev-replace operations on * missing devices when no mirrors are available, but in * this case it should already fail the mount. * This case is handled correctly (but _very_ slowly). */ printk_ratelimited(KERN_WARNING |
efe120a06 Btrfs: convert pr... |
2049 2050 |
"BTRFS: scrub_submit(bio bdev == NULL) is unexpected! "); |
ff023aac3 Btrfs: add code t... |
2051 2052 2053 2054 |
bio_endio(sbio->bio, -EIO); } else { btrfsic_submit_bio(READ, sbio->bio); } |
a2de733c7 btrfs: scrub |
2055 |
} |
ff023aac3 Btrfs: add code t... |
2056 2057 |
static int scrub_add_page_to_rd_bio(struct scrub_ctx *sctx, struct scrub_page *spage) |
a2de733c7 btrfs: scrub |
2058 |
{ |
b5d67f64f Btrfs: change scr... |
2059 |
struct scrub_block *sblock = spage->sblock; |
a2de733c7 btrfs: scrub |
2060 |
struct scrub_bio *sbio; |
69f4cb526 Btrfs: handle bio... |
2061 |
int ret; |
a2de733c7 btrfs: scrub |
2062 2063 2064 2065 2066 |
again: /* * grab a fresh bio or wait for one to become available */ |
d9d181c1b Btrfs: rename the... |
2067 2068 2069 2070 2071 2072 2073 2074 |
while (sctx->curr == -1) { spin_lock(&sctx->list_lock); sctx->curr = sctx->first_free; if (sctx->curr != -1) { sctx->first_free = sctx->bios[sctx->curr]->next_free; sctx->bios[sctx->curr]->next_free = -1; sctx->bios[sctx->curr]->page_count = 0; spin_unlock(&sctx->list_lock); |
a2de733c7 btrfs: scrub |
2075 |
} else { |
d9d181c1b Btrfs: rename the... |
2076 2077 |
spin_unlock(&sctx->list_lock); wait_event(sctx->list_wait, sctx->first_free != -1); |
a2de733c7 btrfs: scrub |
2078 2079 |
} } |
d9d181c1b Btrfs: rename the... |
2080 |
sbio = sctx->bios[sctx->curr]; |
b5d67f64f Btrfs: change scr... |
2081 |
if (sbio->page_count == 0) { |
69f4cb526 Btrfs: handle bio... |
2082 |
struct bio *bio; |
b5d67f64f Btrfs: change scr... |
2083 2084 |
sbio->physical = spage->physical; sbio->logical = spage->logical; |
a36cf8b89 Btrfs: remove the... |
2085 |
sbio->dev = spage->dev; |
b5d67f64f Btrfs: change scr... |
2086 2087 |
bio = sbio->bio; if (!bio) { |
9be3395bc Btrfs: use a btrf... |
2088 |
bio = btrfs_io_bio_alloc(GFP_NOFS, sctx->pages_per_rd_bio); |
b5d67f64f Btrfs: change scr... |
2089 2090 2091 2092 |
if (!bio) return -ENOMEM; sbio->bio = bio; } |
69f4cb526 Btrfs: handle bio... |
2093 2094 2095 |
bio->bi_private = sbio; bio->bi_end_io = scrub_bio_end_io; |
a36cf8b89 Btrfs: remove the... |
2096 |
bio->bi_bdev = sbio->dev->bdev; |
4f024f379 block: Abstract o... |
2097 |
bio->bi_iter.bi_sector = sbio->physical >> 9; |
69f4cb526 Btrfs: handle bio... |
2098 |
sbio->err = 0; |
b5d67f64f Btrfs: change scr... |
2099 2100 2101 |
} else if (sbio->physical + sbio->page_count * PAGE_SIZE != spage->physical || sbio->logical + sbio->page_count * PAGE_SIZE != |
a36cf8b89 Btrfs: remove the... |
2102 2103 |
spage->logical || sbio->dev != spage->dev) { |
d9d181c1b Btrfs: rename the... |
2104 |
scrub_submit(sctx); |
a2de733c7 btrfs: scrub |
2105 2106 |
goto again; } |
69f4cb526 Btrfs: handle bio... |
2107 |
|
b5d67f64f Btrfs: change scr... |
2108 2109 2110 2111 2112 2113 2114 2115 |
sbio->pagev[sbio->page_count] = spage; ret = bio_add_page(sbio->bio, spage->page, PAGE_SIZE, 0); if (ret != PAGE_SIZE) { if (sbio->page_count < 1) { bio_put(sbio->bio); sbio->bio = NULL; return -EIO; } |
d9d181c1b Btrfs: rename the... |
2116 |
scrub_submit(sctx); |
69f4cb526 Btrfs: handle bio... |
2117 2118 |
goto again; } |
ff023aac3 Btrfs: add code t... |
2119 |
scrub_block_get(sblock); /* one for the page added to the bio */ |
b5d67f64f Btrfs: change scr... |
2120 2121 |
atomic_inc(&sblock->outstanding_pages); sbio->page_count++; |
ff023aac3 Btrfs: add code t... |
2122 |
if (sbio->page_count == sctx->pages_per_rd_bio) |
d9d181c1b Btrfs: rename the... |
2123 |
scrub_submit(sctx); |
b5d67f64f Btrfs: change scr... |
2124 2125 2126 |
return 0; } |
d9d181c1b Btrfs: rename the... |
2127 |
static int scrub_pages(struct scrub_ctx *sctx, u64 logical, u64 len, |
a36cf8b89 Btrfs: remove the... |
2128 |
u64 physical, struct btrfs_device *dev, u64 flags, |
ff023aac3 Btrfs: add code t... |
2129 2130 |
u64 gen, int mirror_num, u8 *csum, int force, u64 physical_for_dev_replace) |
b5d67f64f Btrfs: change scr... |
2131 2132 2133 2134 2135 2136 |
{ struct scrub_block *sblock; int index; sblock = kzalloc(sizeof(*sblock), GFP_NOFS); if (!sblock) { |
d9d181c1b Btrfs: rename the... |
2137 2138 2139 |
spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); |
b5d67f64f Btrfs: change scr... |
2140 |
return -ENOMEM; |
a2de733c7 btrfs: scrub |
2141 |
} |
b5d67f64f Btrfs: change scr... |
2142 |
|
7a9e99876 Btrfs: make the s... |
2143 2144 |
/* one ref inside this function, plus one for each page added to * a bio later on */ |
b5d67f64f Btrfs: change scr... |
2145 |
atomic_set(&sblock->ref_count, 1); |
d9d181c1b Btrfs: rename the... |
2146 |
sblock->sctx = sctx; |
b5d67f64f Btrfs: change scr... |
2147 2148 2149 |
sblock->no_io_error_seen = 1; for (index = 0; len > 0; index++) { |
7a9e99876 Btrfs: make the s... |
2150 |
struct scrub_page *spage; |
b5d67f64f Btrfs: change scr... |
2151 |
u64 l = min_t(u64, len, PAGE_SIZE); |
7a9e99876 Btrfs: make the s... |
2152 2153 2154 |
spage = kzalloc(sizeof(*spage), GFP_NOFS); if (!spage) { leave_nomem: |
d9d181c1b Btrfs: rename the... |
2155 2156 2157 |
spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); |
7a9e99876 Btrfs: make the s... |
2158 |
scrub_block_put(sblock); |
b5d67f64f Btrfs: change scr... |
2159 2160 |
return -ENOMEM; } |
7a9e99876 Btrfs: make the s... |
2161 2162 2163 |
BUG_ON(index >= SCRUB_MAX_PAGES_PER_BLOCK); scrub_page_get(spage); sblock->pagev[index] = spage; |
b5d67f64f Btrfs: change scr... |
2164 |
spage->sblock = sblock; |
a36cf8b89 Btrfs: remove the... |
2165 |
spage->dev = dev; |
b5d67f64f Btrfs: change scr... |
2166 2167 2168 2169 |
spage->flags = flags; spage->generation = gen; spage->logical = logical; spage->physical = physical; |
ff023aac3 Btrfs: add code t... |
2170 |
spage->physical_for_dev_replace = physical_for_dev_replace; |
b5d67f64f Btrfs: change scr... |
2171 2172 2173 |
spage->mirror_num = mirror_num; if (csum) { spage->have_csum = 1; |
d9d181c1b Btrfs: rename the... |
2174 |
memcpy(spage->csum, csum, sctx->csum_size); |
b5d67f64f Btrfs: change scr... |
2175 2176 2177 2178 |
} else { spage->have_csum = 0; } sblock->page_count++; |
7a9e99876 Btrfs: make the s... |
2179 2180 2181 |
spage->page = alloc_page(GFP_NOFS); if (!spage->page) goto leave_nomem; |
b5d67f64f Btrfs: change scr... |
2182 2183 2184 |
len -= l; logical += l; physical += l; |
ff023aac3 Btrfs: add code t... |
2185 |
physical_for_dev_replace += l; |
b5d67f64f Btrfs: change scr... |
2186 |
} |
7a9e99876 Btrfs: make the s... |
2187 |
WARN_ON(sblock->page_count == 0); |
b5d67f64f Btrfs: change scr... |
2188 |
for (index = 0; index < sblock->page_count; index++) { |
7a9e99876 Btrfs: make the s... |
2189 |
struct scrub_page *spage = sblock->pagev[index]; |
1bc877934 btrfs: scrub: don... |
2190 |
int ret; |
ff023aac3 Btrfs: add code t... |
2191 |
ret = scrub_add_page_to_rd_bio(sctx, spage); |
b5d67f64f Btrfs: change scr... |
2192 2193 |
if (ret) { scrub_block_put(sblock); |
1bc877934 btrfs: scrub: don... |
2194 |
return ret; |
b5d67f64f Btrfs: change scr... |
2195 |
} |
1bc877934 btrfs: scrub: don... |
2196 |
} |
a2de733c7 btrfs: scrub |
2197 |
|
b5d67f64f Btrfs: change scr... |
2198 |
if (force) |
d9d181c1b Btrfs: rename the... |
2199 |
scrub_submit(sctx); |
a2de733c7 btrfs: scrub |
2200 |
|
b5d67f64f Btrfs: change scr... |
2201 2202 |
/* last one frees, either here or in bio completion for last page */ scrub_block_put(sblock); |
a2de733c7 btrfs: scrub |
2203 2204 |
return 0; } |
b5d67f64f Btrfs: change scr... |
2205 2206 2207 |
static void scrub_bio_end_io(struct bio *bio, int err) { struct scrub_bio *sbio = bio->bi_private; |
a36cf8b89 Btrfs: remove the... |
2208 |
struct btrfs_fs_info *fs_info = sbio->dev->dev_root->fs_info; |
b5d67f64f Btrfs: change scr... |
2209 2210 2211 |
sbio->err = err; sbio->bio = bio; |
0339ef2f4 btrfs: Replace fs... |
2212 |
btrfs_queue_work(fs_info->scrub_workers, &sbio->work); |
b5d67f64f Btrfs: change scr... |
2213 2214 2215 2216 2217 |
} static void scrub_bio_end_io_worker(struct btrfs_work *work) { struct scrub_bio *sbio = container_of(work, struct scrub_bio, work); |
d9d181c1b Btrfs: rename the... |
2218 |
struct scrub_ctx *sctx = sbio->sctx; |
b5d67f64f Btrfs: change scr... |
2219 |
int i; |
ff023aac3 Btrfs: add code t... |
2220 |
BUG_ON(sbio->page_count > SCRUB_PAGES_PER_RD_BIO); |
b5d67f64f Btrfs: change scr... |
2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 |
if (sbio->err) { for (i = 0; i < sbio->page_count; i++) { struct scrub_page *spage = sbio->pagev[i]; spage->io_error = 1; spage->sblock->no_io_error_seen = 0; } } /* now complete the scrub_block items that have all pages completed */ for (i = 0; i < sbio->page_count; i++) { struct scrub_page *spage = sbio->pagev[i]; struct scrub_block *sblock = spage->sblock; if (atomic_dec_and_test(&sblock->outstanding_pages)) scrub_block_complete(sblock); scrub_block_put(sblock); } |
b5d67f64f Btrfs: change scr... |
2239 2240 |
bio_put(sbio->bio); sbio->bio = NULL; |
d9d181c1b Btrfs: rename the... |
2241 2242 2243 2244 |
spin_lock(&sctx->list_lock); sbio->next_free = sctx->first_free; sctx->first_free = sbio->index; spin_unlock(&sctx->list_lock); |
ff023aac3 Btrfs: add code t... |
2245 2246 2247 2248 2249 2250 2251 |
if (sctx->is_dev_replace && atomic_read(&sctx->wr_ctx.flush_all_writes)) { mutex_lock(&sctx->wr_ctx.wr_lock); scrub_wr_submit(sctx); mutex_unlock(&sctx->wr_ctx.wr_lock); } |
b6bfebc13 Btrfs: cleanup sc... |
2252 |
scrub_pending_bio_dec(sctx); |
b5d67f64f Btrfs: change scr... |
2253 |
} |
5a6ac9eac Btrfs, raid56: su... |
2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 |
static inline void __scrub_mark_bitmap(struct scrub_parity *sparity, unsigned long *bitmap, u64 start, u64 len) { int offset; int nsectors; int sectorsize = sparity->sctx->dev_root->sectorsize; if (len >= sparity->stripe_len) { bitmap_set(bitmap, 0, sparity->nsectors); return; } start -= sparity->logic_start; offset = (int)do_div(start, sparity->stripe_len); offset /= sectorsize; nsectors = (int)len / sectorsize; if (offset + nsectors <= sparity->nsectors) { bitmap_set(bitmap, offset, nsectors); return; } bitmap_set(bitmap, offset, sparity->nsectors - offset); bitmap_set(bitmap, 0, nsectors - (sparity->nsectors - offset)); } static inline void scrub_parity_mark_sectors_error(struct scrub_parity *sparity, u64 start, u64 len) { __scrub_mark_bitmap(sparity, sparity->ebitmap, start, len); } static inline void scrub_parity_mark_sectors_data(struct scrub_parity *sparity, u64 start, u64 len) { __scrub_mark_bitmap(sparity, sparity->dbitmap, start, len); } |
b5d67f64f Btrfs: change scr... |
2292 2293 |
static void scrub_block_complete(struct scrub_block *sblock) { |
5a6ac9eac Btrfs, raid56: su... |
2294 |
int corrupted = 0; |
ff023aac3 Btrfs: add code t... |
2295 |
if (!sblock->no_io_error_seen) { |
5a6ac9eac Btrfs, raid56: su... |
2296 |
corrupted = 1; |
b5d67f64f Btrfs: change scr... |
2297 |
scrub_handle_errored_block(sblock); |
ff023aac3 Btrfs: add code t... |
2298 2299 2300 2301 2302 2303 |
} else { /* * if has checksum error, write via repair mechanism in * dev replace case, otherwise write here in dev replace * case. */ |
5a6ac9eac Btrfs, raid56: su... |
2304 2305 |
corrupted = scrub_checksum(sblock); if (!corrupted && sblock->sctx->is_dev_replace) |
ff023aac3 Btrfs: add code t... |
2306 2307 |
scrub_write_block_to_dev_replace(sblock); } |
5a6ac9eac Btrfs, raid56: su... |
2308 2309 2310 2311 2312 2313 2314 2315 2316 |
if (sblock->sparity && corrupted && !sblock->data_corrected) { u64 start = sblock->pagev[0]->logical; u64 end = sblock->pagev[sblock->page_count - 1]->logical + PAGE_SIZE; scrub_parity_mark_sectors_error(sblock->sparity, start, end - start); } |
b5d67f64f Btrfs: change scr... |
2317 |
} |
d9d181c1b Btrfs: rename the... |
2318 |
static int scrub_find_csum(struct scrub_ctx *sctx, u64 logical, u64 len, |
a2de733c7 btrfs: scrub |
2319 2320 2321 |
u8 *csum) { struct btrfs_ordered_sum *sum = NULL; |
f51a4a182 Btrfs: remove btr... |
2322 |
unsigned long index; |
a2de733c7 btrfs: scrub |
2323 |
unsigned long num_sectors; |
a2de733c7 btrfs: scrub |
2324 |
|
d9d181c1b Btrfs: rename the... |
2325 2326 |
while (!list_empty(&sctx->csum_list)) { sum = list_first_entry(&sctx->csum_list, |
a2de733c7 btrfs: scrub |
2327 2328 2329 2330 2331 |
struct btrfs_ordered_sum, list); if (sum->bytenr > logical) return 0; if (sum->bytenr + sum->len > logical) break; |
d9d181c1b Btrfs: rename the... |
2332 |
++sctx->stat.csum_discards; |
a2de733c7 btrfs: scrub |
2333 2334 2335 2336 2337 2338 |
list_del(&sum->list); kfree(sum); sum = NULL; } if (!sum) return 0; |
f51a4a182 Btrfs: remove btr... |
2339 |
index = ((u32)(logical - sum->bytenr)) / sctx->sectorsize; |
d9d181c1b Btrfs: rename the... |
2340 |
num_sectors = sum->len / sctx->sectorsize; |
f51a4a182 Btrfs: remove btr... |
2341 2342 |
memcpy(csum, sum->sums + index, sctx->csum_size); if (index == num_sectors - 1) { |
a2de733c7 btrfs: scrub |
2343 2344 2345 |
list_del(&sum->list); kfree(sum); } |
f51a4a182 Btrfs: remove btr... |
2346 |
return 1; |
a2de733c7 btrfs: scrub |
2347 2348 2349 |
} /* scrub extent tries to collect up to 64 kB for each bio */ |
d9d181c1b Btrfs: rename the... |
2350 |
static int scrub_extent(struct scrub_ctx *sctx, u64 logical, u64 len, |
a36cf8b89 Btrfs: remove the... |
2351 |
u64 physical, struct btrfs_device *dev, u64 flags, |
ff023aac3 Btrfs: add code t... |
2352 |
u64 gen, int mirror_num, u64 physical_for_dev_replace) |
a2de733c7 btrfs: scrub |
2353 2354 2355 |
{ int ret; u8 csum[BTRFS_CSUM_SIZE]; |
b5d67f64f Btrfs: change scr... |
2356 2357 2358 |
u32 blocksize; if (flags & BTRFS_EXTENT_FLAG_DATA) { |
d9d181c1b Btrfs: rename the... |
2359 2360 2361 2362 2363 |
blocksize = sctx->sectorsize; spin_lock(&sctx->stat_lock); sctx->stat.data_extents_scrubbed++; sctx->stat.data_bytes_scrubbed += len; spin_unlock(&sctx->stat_lock); |
b5d67f64f Btrfs: change scr... |
2364 |
} else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { |
d9d181c1b Btrfs: rename the... |
2365 2366 2367 2368 2369 |
blocksize = sctx->nodesize; spin_lock(&sctx->stat_lock); sctx->stat.tree_extents_scrubbed++; sctx->stat.tree_bytes_scrubbed += len; spin_unlock(&sctx->stat_lock); |
b5d67f64f Btrfs: change scr... |
2370 |
} else { |
d9d181c1b Btrfs: rename the... |
2371 |
blocksize = sctx->sectorsize; |
ff023aac3 Btrfs: add code t... |
2372 |
WARN_ON(1); |
b5d67f64f Btrfs: change scr... |
2373 |
} |
a2de733c7 btrfs: scrub |
2374 2375 |
while (len) { |
b5d67f64f Btrfs: change scr... |
2376 |
u64 l = min_t(u64, len, blocksize); |
a2de733c7 btrfs: scrub |
2377 2378 2379 2380 |
int have_csum = 0; if (flags & BTRFS_EXTENT_FLAG_DATA) { /* push csums to sbio */ |
d9d181c1b Btrfs: rename the... |
2381 |
have_csum = scrub_find_csum(sctx, logical, l, csum); |
a2de733c7 btrfs: scrub |
2382 |
if (have_csum == 0) |
d9d181c1b Btrfs: rename the... |
2383 |
++sctx->stat.no_csum; |
ff023aac3 Btrfs: add code t... |
2384 2385 2386 2387 2388 2389 |
if (sctx->is_dev_replace && !have_csum) { ret = copy_nocow_pages(sctx, logical, l, mirror_num, physical_for_dev_replace); goto behind_scrub_pages; } |
a2de733c7 btrfs: scrub |
2390 |
} |
a36cf8b89 Btrfs: remove the... |
2391 |
ret = scrub_pages(sctx, logical, l, physical, dev, flags, gen, |
ff023aac3 Btrfs: add code t... |
2392 2393 2394 |
mirror_num, have_csum ? csum : NULL, 0, physical_for_dev_replace); behind_scrub_pages: |
a2de733c7 btrfs: scrub |
2395 2396 2397 2398 2399 |
if (ret) return ret; len -= l; logical += l; physical += l; |
ff023aac3 Btrfs: add code t... |
2400 |
physical_for_dev_replace += l; |
a2de733c7 btrfs: scrub |
2401 2402 2403 |
} return 0; } |
5a6ac9eac Btrfs, raid56: su... |
2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 |
static int scrub_pages_for_parity(struct scrub_parity *sparity, u64 logical, u64 len, u64 physical, struct btrfs_device *dev, u64 flags, u64 gen, int mirror_num, u8 *csum) { struct scrub_ctx *sctx = sparity->sctx; struct scrub_block *sblock; int index; sblock = kzalloc(sizeof(*sblock), GFP_NOFS); if (!sblock) { spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); return -ENOMEM; } /* one ref inside this function, plus one for each page added to * a bio later on */ atomic_set(&sblock->ref_count, 1); sblock->sctx = sctx; sblock->no_io_error_seen = 1; sblock->sparity = sparity; scrub_parity_get(sparity); for (index = 0; len > 0; index++) { struct scrub_page *spage; u64 l = min_t(u64, len, PAGE_SIZE); spage = kzalloc(sizeof(*spage), GFP_NOFS); if (!spage) { leave_nomem: spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); scrub_block_put(sblock); return -ENOMEM; } BUG_ON(index >= SCRUB_MAX_PAGES_PER_BLOCK); /* For scrub block */ scrub_page_get(spage); sblock->pagev[index] = spage; /* For scrub parity */ scrub_page_get(spage); list_add_tail(&spage->list, &sparity->spages); spage->sblock = sblock; spage->dev = dev; spage->flags = flags; spage->generation = gen; spage->logical = logical; spage->physical = physical; spage->mirror_num = mirror_num; if (csum) { spage->have_csum = 1; memcpy(spage->csum, csum, sctx->csum_size); } else { spage->have_csum = 0; } sblock->page_count++; spage->page = alloc_page(GFP_NOFS); if (!spage->page) goto leave_nomem; len -= l; logical += l; physical += l; } WARN_ON(sblock->page_count == 0); for (index = 0; index < sblock->page_count; index++) { struct scrub_page *spage = sblock->pagev[index]; int ret; ret = scrub_add_page_to_rd_bio(sctx, spage); if (ret) { scrub_block_put(sblock); return ret; } } /* last one frees, either here or in bio completion for last page */ scrub_block_put(sblock); return 0; } static int scrub_extent_for_parity(struct scrub_parity *sparity, u64 logical, u64 len, u64 physical, struct btrfs_device *dev, u64 flags, u64 gen, int mirror_num) { struct scrub_ctx *sctx = sparity->sctx; int ret; u8 csum[BTRFS_CSUM_SIZE]; u32 blocksize; if (flags & BTRFS_EXTENT_FLAG_DATA) { blocksize = sctx->sectorsize; } else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { blocksize = sctx->nodesize; } else { blocksize = sctx->sectorsize; WARN_ON(1); } while (len) { u64 l = min_t(u64, len, blocksize); int have_csum = 0; if (flags & BTRFS_EXTENT_FLAG_DATA) { /* push csums to sbio */ have_csum = scrub_find_csum(sctx, logical, l, csum); if (have_csum == 0) goto skip; } ret = scrub_pages_for_parity(sparity, logical, l, physical, dev, flags, gen, mirror_num, have_csum ? csum : NULL); |
5a6ac9eac Btrfs, raid56: su... |
2520 2521 |
if (ret) return ret; |
6b6d24b38 Btrfs, scrub: uni... |
2522 |
skip: |
5a6ac9eac Btrfs, raid56: su... |
2523 2524 2525 2526 2527 2528 |
len -= l; logical += l; physical += l; } return 0; } |
3b080b256 Btrfs: scrub raid... |
2529 2530 2531 2532 2533 2534 2535 2536 |
/* * Given a physical address, this will calculate it's * logical offset. if this is a parity stripe, it will return * the most left data stripe's logical offset. * * return 0 if it is a data stripe, 1 means parity stripe. */ static int get_raid56_logic_offset(u64 physical, int num, |
5a6ac9eac Btrfs, raid56: su... |
2537 2538 |
struct map_lookup *map, u64 *offset, u64 *stripe_start) |
3b080b256 Btrfs: scrub raid... |
2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 |
{ int i; int j = 0; u64 stripe_nr; u64 last_offset; int stripe_index; int rot; last_offset = (physical - map->stripes[num].physical) * nr_data_stripes(map); |
5a6ac9eac Btrfs, raid56: su... |
2549 2550 |
if (stripe_start) *stripe_start = last_offset; |
3b080b256 Btrfs: scrub raid... |
2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 |
*offset = last_offset; for (i = 0; i < nr_data_stripes(map); i++) { *offset = last_offset + i * map->stripe_len; stripe_nr = *offset; do_div(stripe_nr, map->stripe_len); do_div(stripe_nr, nr_data_stripes(map)); /* Work out the disk rotation on this stripe-set */ rot = do_div(stripe_nr, map->num_stripes); /* calculate which stripe this data locates */ rot += i; |
e4fbaee29 Btrfs: fix compil... |
2563 |
stripe_index = rot % map->num_stripes; |
3b080b256 Btrfs: scrub raid... |
2564 2565 2566 2567 2568 2569 2570 2571 |
if (stripe_index == num) return 0; if (stripe_index < num) j++; } *offset = last_offset + j * map->stripe_len; return 1; } |
5a6ac9eac Btrfs, raid56: su... |
2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 |
static void scrub_free_parity(struct scrub_parity *sparity) { struct scrub_ctx *sctx = sparity->sctx; struct scrub_page *curr, *next; int nbits; nbits = bitmap_weight(sparity->ebitmap, sparity->nsectors); if (nbits) { spin_lock(&sctx->stat_lock); sctx->stat.read_errors += nbits; sctx->stat.uncorrectable_errors += nbits; spin_unlock(&sctx->stat_lock); } list_for_each_entry_safe(curr, next, &sparity->spages, list) { list_del_init(&curr->list); scrub_page_put(curr); } kfree(sparity); } static void scrub_parity_bio_endio(struct bio *bio, int error) { struct scrub_parity *sparity = (struct scrub_parity *)bio->bi_private; struct scrub_ctx *sctx = sparity->sctx; if (error) bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, sparity->nsectors); scrub_free_parity(sparity); scrub_pending_bio_dec(sctx); bio_put(bio); } static void scrub_parity_check_and_repair(struct scrub_parity *sparity) { struct scrub_ctx *sctx = sparity->sctx; struct bio *bio; struct btrfs_raid_bio *rbio; struct scrub_page *spage; struct btrfs_bio *bbio = NULL; u64 *raid_map = NULL; u64 length; int ret; if (!bitmap_andnot(sparity->dbitmap, sparity->dbitmap, sparity->ebitmap, sparity->nsectors)) goto out; length = sparity->logic_end - sparity->logic_start + 1; |
760359769 Btrfs, replace: w... |
2624 |
ret = btrfs_map_sblock(sctx->dev_root->fs_info, WRITE, |
5a6ac9eac Btrfs, raid56: su... |
2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 |
sparity->logic_start, &length, &bbio, 0, &raid_map); if (ret || !bbio || !raid_map) goto bbio_out; bio = btrfs_io_bio_alloc(GFP_NOFS, 0); if (!bio) goto bbio_out; bio->bi_iter.bi_sector = sparity->logic_start >> 9; bio->bi_private = sparity; bio->bi_end_io = scrub_parity_bio_endio; rbio = raid56_parity_alloc_scrub_rbio(sctx->dev_root, bio, bbio, raid_map, length, sparity->scrub_dev, sparity->dbitmap, sparity->nsectors); if (!rbio) goto rbio_out; list_for_each_entry(spage, &sparity->spages, list) raid56_parity_add_scrub_pages(rbio, spage->page, spage->logical); scrub_pending_bio_inc(sctx); raid56_parity_submit_scrub_rbio(rbio); return; rbio_out: bio_put(bio); bbio_out: kfree(bbio); kfree(raid_map); bitmap_or(sparity->ebitmap, sparity->ebitmap, sparity->dbitmap, sparity->nsectors); spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); out: scrub_free_parity(sparity); } static inline int scrub_calc_parity_bitmap_len(int nsectors) { return DIV_ROUND_UP(nsectors, BITS_PER_LONG) * (BITS_PER_LONG / 8); } static void scrub_parity_get(struct scrub_parity *sparity) { atomic_inc(&sparity->ref_count); } static void scrub_parity_put(struct scrub_parity *sparity) { if (!atomic_dec_and_test(&sparity->ref_count)) return; scrub_parity_check_and_repair(sparity); } static noinline_for_stack int scrub_raid56_parity(struct scrub_ctx *sctx, struct map_lookup *map, struct btrfs_device *sdev, struct btrfs_path *path, u64 logic_start, u64 logic_end) { struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; struct btrfs_root *root = fs_info->extent_root; struct btrfs_root *csum_root = fs_info->csum_root; struct btrfs_extent_item *extent; u64 flags; int ret; int slot; struct extent_buffer *l; struct btrfs_key key; u64 generation; u64 extent_logical; u64 extent_physical; u64 extent_len; struct btrfs_device *extent_dev; struct scrub_parity *sparity; int nsectors; int bitmap_len; int extent_mirror_num; int stop_loop = 0; nsectors = map->stripe_len / root->sectorsize; bitmap_len = scrub_calc_parity_bitmap_len(nsectors); sparity = kzalloc(sizeof(struct scrub_parity) + 2 * bitmap_len, GFP_NOFS); if (!sparity) { spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); return -ENOMEM; } sparity->stripe_len = map->stripe_len; sparity->nsectors = nsectors; sparity->sctx = sctx; sparity->scrub_dev = sdev; sparity->logic_start = logic_start; sparity->logic_end = logic_end; atomic_set(&sparity->ref_count, 1); INIT_LIST_HEAD(&sparity->spages); sparity->dbitmap = sparity->bitmap; sparity->ebitmap = (void *)sparity->bitmap + bitmap_len; ret = 0; while (logic_start < logic_end) { if (btrfs_fs_incompat(fs_info, SKINNY_METADATA)) key.type = BTRFS_METADATA_ITEM_KEY; else key.type = BTRFS_EXTENT_ITEM_KEY; key.objectid = logic_start; key.offset = (u64)-1; ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto out; if (ret > 0) { ret = btrfs_previous_extent_item(root, path, 0); if (ret < 0) goto out; if (ret > 0) { btrfs_release_path(path); ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto out; } } stop_loop = 0; while (1) { u64 bytes; l = path->nodes[0]; slot = path->slots[0]; if (slot >= btrfs_header_nritems(l)) { ret = btrfs_next_leaf(root, path); if (ret == 0) continue; if (ret < 0) goto out; stop_loop = 1; break; } btrfs_item_key_to_cpu(l, &key, slot); if (key.type == BTRFS_METADATA_ITEM_KEY) bytes = root->nodesize; else bytes = key.offset; if (key.objectid + bytes <= logic_start) goto next; if (key.type != BTRFS_EXTENT_ITEM_KEY && key.type != BTRFS_METADATA_ITEM_KEY) goto next; if (key.objectid > logic_end) { stop_loop = 1; break; } while (key.objectid >= logic_start + map->stripe_len) logic_start += map->stripe_len; extent = btrfs_item_ptr(l, slot, struct btrfs_extent_item); flags = btrfs_extent_flags(l, extent); generation = btrfs_extent_generation(l, extent); if (key.objectid < logic_start && (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) { btrfs_err(fs_info, "scrub: tree block %llu spanning stripes, ignored. logical=%llu", key.objectid, logic_start); goto next; } again: extent_logical = key.objectid; extent_len = bytes; if (extent_logical < logic_start) { extent_len -= logic_start - extent_logical; extent_logical = logic_start; } if (extent_logical + extent_len > logic_start + map->stripe_len) extent_len = logic_start + map->stripe_len - extent_logical; scrub_parity_mark_sectors_data(sparity, extent_logical, extent_len); scrub_remap_extent(fs_info, extent_logical, extent_len, &extent_physical, &extent_dev, &extent_mirror_num); ret = btrfs_lookup_csums_range(csum_root, extent_logical, extent_logical + extent_len - 1, &sctx->csum_list, 1); if (ret) goto out; ret = scrub_extent_for_parity(sparity, extent_logical, extent_len, extent_physical, extent_dev, flags, generation, extent_mirror_num); if (ret) goto out; scrub_free_csums(sctx); if (extent_logical + extent_len < key.objectid + bytes) { logic_start += map->stripe_len; if (logic_start >= logic_end) { stop_loop = 1; break; } if (logic_start < key.objectid + bytes) { cond_resched(); goto again; } } next: path->slots[0]++; } btrfs_release_path(path); if (stop_loop) break; logic_start += map->stripe_len; } out: if (ret < 0) scrub_parity_mark_sectors_error(sparity, logic_start, logic_end - logic_start + 1); scrub_parity_put(sparity); scrub_submit(sctx); mutex_lock(&sctx->wr_ctx.wr_lock); scrub_wr_submit(sctx); mutex_unlock(&sctx->wr_ctx.wr_lock); btrfs_release_path(path); return ret < 0 ? ret : 0; } |
d9d181c1b Btrfs: rename the... |
2888 |
static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx, |
a36cf8b89 Btrfs: remove the... |
2889 2890 |
struct map_lookup *map, struct btrfs_device *scrub_dev, |
ff023aac3 Btrfs: add code t... |
2891 2892 |
int num, u64 base, u64 length, int is_dev_replace) |
a2de733c7 btrfs: scrub |
2893 |
{ |
5a6ac9eac Btrfs, raid56: su... |
2894 |
struct btrfs_path *path, *ppath; |
a36cf8b89 Btrfs: remove the... |
2895 |
struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; |
a2de733c7 btrfs: scrub |
2896 2897 2898 |
struct btrfs_root *root = fs_info->extent_root; struct btrfs_root *csum_root = fs_info->csum_root; struct btrfs_extent_item *extent; |
e7786c3ae btrfs: scrub: add... |
2899 |
struct blk_plug plug; |
a2de733c7 btrfs: scrub |
2900 2901 2902 |
u64 flags; int ret; int slot; |
a2de733c7 btrfs: scrub |
2903 |
u64 nstripes; |
a2de733c7 btrfs: scrub |
2904 2905 2906 2907 |
struct extent_buffer *l; struct btrfs_key key; u64 physical; u64 logical; |
625f1c8dc Btrfs: improve th... |
2908 |
u64 logic_end; |
3b080b256 Btrfs: scrub raid... |
2909 |
u64 physical_end; |
a2de733c7 btrfs: scrub |
2910 |
u64 generation; |
e12fa9cd3 btrfs scrub: use ... |
2911 |
int mirror_num; |
7a26285ee btrfs: use readah... |
2912 2913 2914 2915 |
struct reada_control *reada1; struct reada_control *reada2; struct btrfs_key key_start; struct btrfs_key key_end; |
a2de733c7 btrfs: scrub |
2916 2917 |
u64 increment = map->stripe_len; u64 offset; |
ff023aac3 Btrfs: add code t... |
2918 2919 2920 |
u64 extent_logical; u64 extent_physical; u64 extent_len; |
5a6ac9eac Btrfs, raid56: su... |
2921 2922 |
u64 stripe_logical; u64 stripe_end; |
ff023aac3 Btrfs: add code t... |
2923 2924 |
struct btrfs_device *extent_dev; int extent_mirror_num; |
3b080b256 Btrfs: scrub raid... |
2925 |
int stop_loop = 0; |
53b381b3a Btrfs: RAID5 and ... |
2926 |
|
a2de733c7 btrfs: scrub |
2927 |
nstripes = length; |
3b080b256 Btrfs: scrub raid... |
2928 |
physical = map->stripes[num].physical; |
a2de733c7 btrfs: scrub |
2929 2930 2931 2932 2933 |
offset = 0; do_div(nstripes, map->stripe_len); if (map->type & BTRFS_BLOCK_GROUP_RAID0) { offset = map->stripe_len * num; increment = map->stripe_len * map->num_stripes; |
193ea74b2 btrfs scrub: bugf... |
2934 |
mirror_num = 1; |
a2de733c7 btrfs: scrub |
2935 2936 2937 2938 |
} else if (map->type & BTRFS_BLOCK_GROUP_RAID10) { int factor = map->num_stripes / map->sub_stripes; offset = map->stripe_len * (num / map->sub_stripes); increment = map->stripe_len * factor; |
193ea74b2 btrfs scrub: bugf... |
2939 |
mirror_num = num % map->sub_stripes + 1; |
a2de733c7 btrfs: scrub |
2940 2941 |
} else if (map->type & BTRFS_BLOCK_GROUP_RAID1) { increment = map->stripe_len; |
193ea74b2 btrfs scrub: bugf... |
2942 |
mirror_num = num % map->num_stripes + 1; |
a2de733c7 btrfs: scrub |
2943 2944 |
} else if (map->type & BTRFS_BLOCK_GROUP_DUP) { increment = map->stripe_len; |
193ea74b2 btrfs scrub: bugf... |
2945 |
mirror_num = num % map->num_stripes + 1; |
3b080b256 Btrfs: scrub raid... |
2946 2947 |
} else if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { |
5a6ac9eac Btrfs, raid56: su... |
2948 |
get_raid56_logic_offset(physical, num, map, &offset, NULL); |
3b080b256 Btrfs: scrub raid... |
2949 2950 |
increment = map->stripe_len * nr_data_stripes(map); mirror_num = 1; |
a2de733c7 btrfs: scrub |
2951 2952 |
} else { increment = map->stripe_len; |
193ea74b2 btrfs scrub: bugf... |
2953 |
mirror_num = 1; |
a2de733c7 btrfs: scrub |
2954 2955 2956 2957 2958 |
} path = btrfs_alloc_path(); if (!path) return -ENOMEM; |
5a6ac9eac Btrfs, raid56: su... |
2959 2960 2961 2962 2963 |
ppath = btrfs_alloc_path(); if (!ppath) { btrfs_free_path(ppath); return -ENOMEM; } |
b5d67f64f Btrfs: change scr... |
2964 2965 2966 2967 2968 |
/* * work on commit root. The related disk blocks are static as * long as COW is applied. This means, it is save to rewrite * them to repair disk errors without any race conditions */ |
a2de733c7 btrfs: scrub |
2969 2970 2971 2972 |
path->search_commit_root = 1; path->skip_locking = 1; /* |
7a26285ee btrfs: use readah... |
2973 2974 2975 |
* trigger the readahead for extent tree csum tree and wait for * completion. During readahead, the scrub is officially paused * to not hold off transaction commits |
a2de733c7 btrfs: scrub |
2976 2977 |
*/ logical = base + offset; |
3b080b256 Btrfs: scrub raid... |
2978 2979 2980 2981 |
physical_end = physical + nstripes * map->stripe_len; if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { get_raid56_logic_offset(physical_end, num, |
5a6ac9eac Btrfs, raid56: su... |
2982 |
map, &logic_end, NULL); |
3b080b256 Btrfs: scrub raid... |
2983 2984 2985 2986 |
logic_end += base; } else { logic_end = logical + increment * nstripes; } |
d9d181c1b Btrfs: rename the... |
2987 |
wait_event(sctx->list_wait, |
b6bfebc13 Btrfs: cleanup sc... |
2988 |
atomic_read(&sctx->bios_in_flight) == 0); |
cb7ab0215 Btrfs: wrap repea... |
2989 |
scrub_blocked_if_needed(fs_info); |
7a26285ee btrfs: use readah... |
2990 2991 2992 2993 2994 |
/* FIXME it might be better to start readahead at commit root */ key_start.objectid = logical; key_start.type = BTRFS_EXTENT_ITEM_KEY; key_start.offset = (u64)0; |
3b080b256 Btrfs: scrub raid... |
2995 |
key_end.objectid = logic_end; |
3173a18f7 Btrfs: add a inco... |
2996 2997 |
key_end.type = BTRFS_METADATA_ITEM_KEY; key_end.offset = (u64)-1; |
7a26285ee btrfs: use readah... |
2998 2999 3000 3001 3002 3003 3004 |
reada1 = btrfs_reada_add(root, &key_start, &key_end); key_start.objectid = BTRFS_EXTENT_CSUM_OBJECTID; key_start.type = BTRFS_EXTENT_CSUM_KEY; key_start.offset = logical; key_end.objectid = BTRFS_EXTENT_CSUM_OBJECTID; key_end.type = BTRFS_EXTENT_CSUM_KEY; |
3b080b256 Btrfs: scrub raid... |
3005 |
key_end.offset = logic_end; |
7a26285ee btrfs: use readah... |
3006 3007 3008 3009 3010 3011 |
reada2 = btrfs_reada_add(csum_root, &key_start, &key_end); if (!IS_ERR(reada1)) btrfs_reada_wait(reada1); if (!IS_ERR(reada2)) btrfs_reada_wait(reada2); |
a2de733c7 btrfs: scrub |
3012 3013 3014 3015 3016 |
/* * collect all data csums for the stripe to avoid seeking during * the scrub. This might currently (crc32) end up to be about 1MB */ |
e7786c3ae btrfs: scrub: add... |
3017 |
blk_start_plug(&plug); |
a2de733c7 btrfs: scrub |
3018 |
|
a2de733c7 btrfs: scrub |
3019 3020 3021 |
/* * now find all extents for each stripe and scrub them */ |
a2de733c7 btrfs: scrub |
3022 |
ret = 0; |
3b080b256 Btrfs: scrub raid... |
3023 3024 3025 3026 3027 |
while (physical < physical_end) { /* for raid56, we skip parity stripe */ if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { ret = get_raid56_logic_offset(physical, num, |
5a6ac9eac Btrfs, raid56: su... |
3028 |
map, &logical, &stripe_logical); |
3b080b256 Btrfs: scrub raid... |
3029 |
logical += base; |
5a6ac9eac Btrfs, raid56: su... |
3030 3031 3032 3033 3034 3035 3036 3037 |
if (ret) { stripe_logical += base; stripe_end = stripe_logical + increment - 1; ret = scrub_raid56_parity(sctx, map, scrub_dev, ppath, stripe_logical, stripe_end); if (ret) goto out; |
3b080b256 Btrfs: scrub raid... |
3038 |
goto skip; |
5a6ac9eac Btrfs, raid56: su... |
3039 |
} |
3b080b256 Btrfs: scrub raid... |
3040 |
} |
a2de733c7 btrfs: scrub |
3041 3042 3043 3044 |
/* * canceled? */ if (atomic_read(&fs_info->scrub_cancel_req) || |
d9d181c1b Btrfs: rename the... |
3045 |
atomic_read(&sctx->cancel_req)) { |
a2de733c7 btrfs: scrub |
3046 3047 3048 3049 3050 3051 3052 3053 |
ret = -ECANCELED; goto out; } /* * check to see if we have to pause */ if (atomic_read(&fs_info->scrub_pause_req)) { /* push queued extents */ |
ff023aac3 Btrfs: add code t... |
3054 |
atomic_set(&sctx->wr_ctx.flush_all_writes, 1); |
d9d181c1b Btrfs: rename the... |
3055 |
scrub_submit(sctx); |
ff023aac3 Btrfs: add code t... |
3056 3057 3058 |
mutex_lock(&sctx->wr_ctx.wr_lock); scrub_wr_submit(sctx); mutex_unlock(&sctx->wr_ctx.wr_lock); |
d9d181c1b Btrfs: rename the... |
3059 |
wait_event(sctx->list_wait, |
b6bfebc13 Btrfs: cleanup sc... |
3060 |
atomic_read(&sctx->bios_in_flight) == 0); |
ff023aac3 Btrfs: add code t... |
3061 |
atomic_set(&sctx->wr_ctx.flush_all_writes, 0); |
3cb0929ad Btrfs: fix wrong ... |
3062 |
scrub_blocked_if_needed(fs_info); |
a2de733c7 btrfs: scrub |
3063 |
} |
7c76edb77 Btrfs: fix missin... |
3064 3065 3066 3067 |
if (btrfs_fs_incompat(fs_info, SKINNY_METADATA)) key.type = BTRFS_METADATA_ITEM_KEY; else key.type = BTRFS_EXTENT_ITEM_KEY; |
a2de733c7 btrfs: scrub |
3068 |
key.objectid = logical; |
625f1c8dc Btrfs: improve th... |
3069 |
key.offset = (u64)-1; |
a2de733c7 btrfs: scrub |
3070 3071 3072 3073 |
ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto out; |
3173a18f7 Btrfs: add a inco... |
3074 |
|
8c51032f9 btrfs: scrub: err... |
3075 |
if (ret > 0) { |
ade2e0b3e Btrfs: fix to sea... |
3076 |
ret = btrfs_previous_extent_item(root, path, 0); |
a2de733c7 btrfs: scrub |
3077 3078 |
if (ret < 0) goto out; |
8c51032f9 btrfs: scrub: err... |
3079 3080 3081 3082 3083 3084 3085 3086 3087 |
if (ret > 0) { /* there's no smaller item, so stick with the * larger one */ btrfs_release_path(path); ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto out; } |
a2de733c7 btrfs: scrub |
3088 |
} |
625f1c8dc Btrfs: improve th... |
3089 |
stop_loop = 0; |
a2de733c7 btrfs: scrub |
3090 |
while (1) { |
3173a18f7 Btrfs: add a inco... |
3091 |
u64 bytes; |
a2de733c7 btrfs: scrub |
3092 3093 3094 3095 3096 3097 3098 3099 |
l = path->nodes[0]; slot = path->slots[0]; if (slot >= btrfs_header_nritems(l)) { ret = btrfs_next_leaf(root, path); if (ret == 0) continue; if (ret < 0) goto out; |
625f1c8dc Btrfs: improve th... |
3100 |
stop_loop = 1; |
a2de733c7 btrfs: scrub |
3101 3102 3103 |
break; } btrfs_item_key_to_cpu(l, &key, slot); |
3173a18f7 Btrfs: add a inco... |
3104 |
if (key.type == BTRFS_METADATA_ITEM_KEY) |
707e8a071 btrfs: use nodesi... |
3105 |
bytes = root->nodesize; |
3173a18f7 Btrfs: add a inco... |
3106 3107 3108 3109 |
else bytes = key.offset; if (key.objectid + bytes <= logical) |
a2de733c7 btrfs: scrub |
3110 |
goto next; |
625f1c8dc Btrfs: improve th... |
3111 3112 3113 |
if (key.type != BTRFS_EXTENT_ITEM_KEY && key.type != BTRFS_METADATA_ITEM_KEY) goto next; |
a2de733c7 btrfs: scrub |
3114 |
|
625f1c8dc Btrfs: improve th... |
3115 3116 3117 3118 3119 3120 |
if (key.objectid >= logical + map->stripe_len) { /* out of this device extent */ if (key.objectid >= logic_end) stop_loop = 1; break; } |
a2de733c7 btrfs: scrub |
3121 3122 3123 3124 3125 3126 3127 3128 |
extent = btrfs_item_ptr(l, slot, struct btrfs_extent_item); flags = btrfs_extent_flags(l, extent); generation = btrfs_extent_generation(l, extent); if (key.objectid < logical && (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) { |
efe120a06 Btrfs: convert pr... |
3129 3130 3131 |
btrfs_err(fs_info, "scrub: tree block %llu spanning " "stripes, ignored. logical=%llu", |
c1c9ff7c9 Btrfs: Remove sup... |
3132 |
key.objectid, logical); |
a2de733c7 btrfs: scrub |
3133 3134 |
goto next; } |
625f1c8dc Btrfs: improve th... |
3135 3136 3137 |
again: extent_logical = key.objectid; extent_len = bytes; |
a2de733c7 btrfs: scrub |
3138 3139 3140 |
/* * trim extent to this stripe */ |
625f1c8dc Btrfs: improve th... |
3141 3142 3143 |
if (extent_logical < logical) { extent_len -= logical - extent_logical; extent_logical = logical; |
a2de733c7 btrfs: scrub |
3144 |
} |
625f1c8dc Btrfs: improve th... |
3145 |
if (extent_logical + extent_len > |
a2de733c7 btrfs: scrub |
3146 |
logical + map->stripe_len) { |
625f1c8dc Btrfs: improve th... |
3147 3148 |
extent_len = logical + map->stripe_len - extent_logical; |
a2de733c7 btrfs: scrub |
3149 |
} |
625f1c8dc Btrfs: improve th... |
3150 |
extent_physical = extent_logical - logical + physical; |
ff023aac3 Btrfs: add code t... |
3151 3152 3153 3154 3155 3156 3157 |
extent_dev = scrub_dev; extent_mirror_num = mirror_num; if (is_dev_replace) scrub_remap_extent(fs_info, extent_logical, extent_len, &extent_physical, &extent_dev, &extent_mirror_num); |
625f1c8dc Btrfs: improve th... |
3158 3159 3160 3161 3162 3163 |
ret = btrfs_lookup_csums_range(csum_root, logical, logical + map->stripe_len - 1, &sctx->csum_list, 1); if (ret) goto out; |
ff023aac3 Btrfs: add code t... |
3164 3165 3166 |
ret = scrub_extent(sctx, extent_logical, extent_len, extent_physical, extent_dev, flags, generation, extent_mirror_num, |
115930cb2 Btrfs: fix wrong ... |
3167 |
extent_logical - logical + physical); |
a2de733c7 btrfs: scrub |
3168 3169 |
if (ret) goto out; |
d88d46c6e Btrfs: free csums... |
3170 |
scrub_free_csums(sctx); |
625f1c8dc Btrfs: improve th... |
3171 3172 |
if (extent_logical + extent_len < key.objectid + bytes) { |
3b080b256 Btrfs: scrub raid... |
3173 3174 3175 3176 3177 3178 |
if (map->type & (BTRFS_BLOCK_GROUP_RAID5 | BTRFS_BLOCK_GROUP_RAID6)) { /* * loop until we find next data stripe * or we have finished all stripes. */ |
5a6ac9eac Btrfs, raid56: su... |
3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 |
loop: physical += map->stripe_len; ret = get_raid56_logic_offset(physical, num, map, &logical, &stripe_logical); logical += base; if (ret && physical < physical_end) { stripe_logical += base; stripe_end = stripe_logical + increment - 1; ret = scrub_raid56_parity(sctx, map, scrub_dev, ppath, stripe_logical, stripe_end); if (ret) goto out; goto loop; } |
3b080b256 Btrfs: scrub raid... |
3198 3199 3200 3201 |
} else { physical += map->stripe_len; logical += increment; } |
625f1c8dc Btrfs: improve th... |
3202 3203 3204 3205 |
if (logical < key.objectid + bytes) { cond_resched(); goto again; } |
3b080b256 Btrfs: scrub raid... |
3206 |
if (physical >= physical_end) { |
625f1c8dc Btrfs: improve th... |
3207 3208 3209 3210 |
stop_loop = 1; break; } } |
a2de733c7 btrfs: scrub |
3211 3212 3213 |
next: path->slots[0]++; } |
712673339 Merge branch 'for... |
3214 |
btrfs_release_path(path); |
3b080b256 Btrfs: scrub raid... |
3215 |
skip: |
a2de733c7 btrfs: scrub |
3216 3217 |
logical += increment; physical += map->stripe_len; |
d9d181c1b Btrfs: rename the... |
3218 |
spin_lock(&sctx->stat_lock); |
625f1c8dc Btrfs: improve th... |
3219 3220 3221 3222 3223 |
if (stop_loop) sctx->stat.last_physical = map->stripes[num].physical + length; else sctx->stat.last_physical = physical; |
d9d181c1b Btrfs: rename the... |
3224 |
spin_unlock(&sctx->stat_lock); |
625f1c8dc Btrfs: improve th... |
3225 3226 |
if (stop_loop) break; |
a2de733c7 btrfs: scrub |
3227 |
} |
ff023aac3 Btrfs: add code t... |
3228 |
out: |
a2de733c7 btrfs: scrub |
3229 |
/* push queued extents */ |
d9d181c1b Btrfs: rename the... |
3230 |
scrub_submit(sctx); |
ff023aac3 Btrfs: add code t... |
3231 3232 3233 |
mutex_lock(&sctx->wr_ctx.wr_lock); scrub_wr_submit(sctx); mutex_unlock(&sctx->wr_ctx.wr_lock); |
a2de733c7 btrfs: scrub |
3234 |
|
e7786c3ae btrfs: scrub: add... |
3235 |
blk_finish_plug(&plug); |
a2de733c7 btrfs: scrub |
3236 |
btrfs_free_path(path); |
5a6ac9eac Btrfs, raid56: su... |
3237 |
btrfs_free_path(ppath); |
a2de733c7 btrfs: scrub |
3238 3239 |
return ret < 0 ? ret : 0; } |
d9d181c1b Btrfs: rename the... |
3240 |
static noinline_for_stack int scrub_chunk(struct scrub_ctx *sctx, |
a36cf8b89 Btrfs: remove the... |
3241 3242 3243 |
struct btrfs_device *scrub_dev, u64 chunk_tree, u64 chunk_objectid, u64 chunk_offset, u64 length, |
ff023aac3 Btrfs: add code t... |
3244 |
u64 dev_offset, int is_dev_replace) |
a2de733c7 btrfs: scrub |
3245 3246 |
{ struct btrfs_mapping_tree *map_tree = |
a36cf8b89 Btrfs: remove the... |
3247 |
&sctx->dev_root->fs_info->mapping_tree; |
a2de733c7 btrfs: scrub |
3248 3249 3250 |
struct map_lookup *map; struct extent_map *em; int i; |
ff023aac3 Btrfs: add code t... |
3251 |
int ret = 0; |
a2de733c7 btrfs: scrub |
3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 |
read_lock(&map_tree->map_tree.lock); em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1); read_unlock(&map_tree->map_tree.lock); if (!em) return -EINVAL; map = (struct map_lookup *)em->bdev; if (em->start != chunk_offset) goto out; if (em->len < length) goto out; for (i = 0; i < map->num_stripes; ++i) { |
a36cf8b89 Btrfs: remove the... |
3268 |
if (map->stripes[i].dev->bdev == scrub_dev->bdev && |
859acaf1a btrfs: don't chec... |
3269 |
map->stripes[i].physical == dev_offset) { |
a36cf8b89 Btrfs: remove the... |
3270 |
ret = scrub_stripe(sctx, map, scrub_dev, i, |
ff023aac3 Btrfs: add code t... |
3271 3272 |
chunk_offset, length, is_dev_replace); |
a2de733c7 btrfs: scrub |
3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 |
if (ret) goto out; } } out: free_extent_map(em); return ret; } static noinline_for_stack |
a36cf8b89 Btrfs: remove the... |
3284 |
int scrub_enumerate_chunks(struct scrub_ctx *sctx, |
ff023aac3 Btrfs: add code t... |
3285 3286 |
struct btrfs_device *scrub_dev, u64 start, u64 end, int is_dev_replace) |
a2de733c7 btrfs: scrub |
3287 3288 3289 |
{ struct btrfs_dev_extent *dev_extent = NULL; struct btrfs_path *path; |
a36cf8b89 Btrfs: remove the... |
3290 |
struct btrfs_root *root = sctx->dev_root; |
a2de733c7 btrfs: scrub |
3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 |
struct btrfs_fs_info *fs_info = root->fs_info; u64 length; u64 chunk_tree; u64 chunk_objectid; u64 chunk_offset; int ret; int slot; struct extent_buffer *l; struct btrfs_key key; struct btrfs_key found_key; struct btrfs_block_group_cache *cache; |
ff023aac3 Btrfs: add code t... |
3302 |
struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace; |
a2de733c7 btrfs: scrub |
3303 3304 3305 3306 3307 3308 3309 3310 |
path = btrfs_alloc_path(); if (!path) return -ENOMEM; path->reada = 2; path->search_commit_root = 1; path->skip_locking = 1; |
a36cf8b89 Btrfs: remove the... |
3311 |
key.objectid = scrub_dev->devid; |
a2de733c7 btrfs: scrub |
3312 3313 |
key.offset = 0ull; key.type = BTRFS_DEV_EXTENT_KEY; |
a2de733c7 btrfs: scrub |
3314 3315 3316 |
while (1) { ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) |
8c51032f9 btrfs: scrub: err... |
3317 3318 3319 3320 3321 3322 3323 3324 3325 |
break; if (ret > 0) { if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { ret = btrfs_next_leaf(root, path); if (ret) break; } } |
a2de733c7 btrfs: scrub |
3326 3327 3328 3329 3330 |
l = path->nodes[0]; slot = path->slots[0]; btrfs_item_key_to_cpu(l, &found_key, slot); |
a36cf8b89 Btrfs: remove the... |
3331 |
if (found_key.objectid != scrub_dev->devid) |
a2de733c7 btrfs: scrub |
3332 |
break; |
962a298f3 btrfs: kill the k... |
3333 |
if (found_key.type != BTRFS_DEV_EXTENT_KEY) |
a2de733c7 btrfs: scrub |
3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 |
break; if (found_key.offset >= end) break; if (found_key.offset < key.offset) break; dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent); length = btrfs_dev_extent_length(l, dev_extent); |
ced96edc4 btrfs: Skip scrub... |
3344 3345 |
if (found_key.offset + length <= start) goto skip; |
a2de733c7 btrfs: scrub |
3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 |
chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent); chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent); chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent); /* * get a reference on the corresponding block group to prevent * the chunk from going away while we scrub it */ cache = btrfs_lookup_block_group(fs_info, chunk_offset); |
ced96edc4 btrfs: Skip scrub... |
3356 3357 3358 3359 3360 |
/* some chunks are removed but not committed to disk yet, * continue scrubbing */ if (!cache) goto skip; |
ff023aac3 Btrfs: add code t... |
3361 3362 3363 |
dev_replace->cursor_right = found_key.offset + length; dev_replace->cursor_left = found_key.offset; dev_replace->item_needs_writeback = 1; |
a36cf8b89 Btrfs: remove the... |
3364 |
ret = scrub_chunk(sctx, scrub_dev, chunk_tree, chunk_objectid, |
ff023aac3 Btrfs: add code t... |
3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 |
chunk_offset, length, found_key.offset, is_dev_replace); /* * flush, submit all pending read and write bios, afterwards * wait for them. * Note that in the dev replace case, a read request causes * write requests that are submitted in the read completion * worker. Therefore in the current situation, it is required * that all write requests are flushed, so that all read and * write requests are really completed when bios_in_flight * changes to 0. */ atomic_set(&sctx->wr_ctx.flush_all_writes, 1); scrub_submit(sctx); mutex_lock(&sctx->wr_ctx.wr_lock); scrub_wr_submit(sctx); mutex_unlock(&sctx->wr_ctx.wr_lock); wait_event(sctx->list_wait, atomic_read(&sctx->bios_in_flight) == 0); |
12cf93728 Btrfs: device_rep... |
3386 3387 3388 3389 3390 3391 3392 3393 |
atomic_inc(&fs_info->scrubs_paused); wake_up(&fs_info->scrub_pause_wait); /* * must be called before we decrease @scrub_paused. * make sure we don't block transaction commit while * we are waiting pending workers finished. */ |
ff023aac3 Btrfs: add code t... |
3394 3395 |
wait_event(sctx->list_wait, atomic_read(&sctx->workers_pending) == 0); |
12cf93728 Btrfs: device_rep... |
3396 3397 3398 3399 3400 3401 3402 |
atomic_set(&sctx->wr_ctx.flush_all_writes, 0); mutex_lock(&fs_info->scrub_lock); __scrub_blocked_if_needed(fs_info); atomic_dec(&fs_info->scrubs_paused); mutex_unlock(&fs_info->scrub_lock); wake_up(&fs_info->scrub_pause_wait); |
ff023aac3 Btrfs: add code t... |
3403 |
|
a2de733c7 btrfs: scrub |
3404 3405 3406 |
btrfs_put_block_group(cache); if (ret) break; |
af1be4f85 Btrfs: fix a scru... |
3407 3408 |
if (is_dev_replace && atomic64_read(&dev_replace->num_write_errors) > 0) { |
ff023aac3 Btrfs: add code t... |
3409 3410 3411 3412 3413 3414 3415 |
ret = -EIO; break; } if (sctx->stat.malloc_errors > 0) { ret = -ENOMEM; break; } |
a2de733c7 btrfs: scrub |
3416 |
|
539f358a3 Btrfs: fix the de... |
3417 3418 |
dev_replace->cursor_left = dev_replace->cursor_right; dev_replace->item_needs_writeback = 1; |
ced96edc4 btrfs: Skip scrub... |
3419 |
skip: |
a2de733c7 btrfs: scrub |
3420 |
key.offset = found_key.offset + length; |
712673339 Merge branch 'for... |
3421 |
btrfs_release_path(path); |
a2de733c7 btrfs: scrub |
3422 |
} |
a2de733c7 btrfs: scrub |
3423 |
btrfs_free_path(path); |
8c51032f9 btrfs: scrub: err... |
3424 3425 3426 3427 3428 3429 |
/* * ret can still be 1 from search_slot or next_leaf, * that's not an error */ return ret < 0 ? ret : 0; |
a2de733c7 btrfs: scrub |
3430 |
} |
a36cf8b89 Btrfs: remove the... |
3431 3432 |
static noinline_for_stack int scrub_supers(struct scrub_ctx *sctx, struct btrfs_device *scrub_dev) |
a2de733c7 btrfs: scrub |
3433 3434 3435 3436 3437 |
{ int i; u64 bytenr; u64 gen; int ret; |
a36cf8b89 Btrfs: remove the... |
3438 |
struct btrfs_root *root = sctx->dev_root; |
a2de733c7 btrfs: scrub |
3439 |
|
87533c475 Btrfs: use bit op... |
3440 |
if (test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state)) |
79787eaab btrfs: replace ma... |
3441 |
return -EIO; |
5f546063c Btrfs: fix wrong ... |
3442 3443 3444 3445 3446 |
/* Seed devices of a new filesystem has their own generation. */ if (scrub_dev->fs_devices != root->fs_info->fs_devices) gen = scrub_dev->generation; else gen = root->fs_info->last_trans_committed; |
a2de733c7 btrfs: scrub |
3447 3448 3449 |
for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { bytenr = btrfs_sb_offset(i); |
935e5cc93 Btrfs: fix wrong ... |
3450 3451 |
if (bytenr + BTRFS_SUPER_INFO_SIZE > scrub_dev->commit_total_bytes) |
a2de733c7 btrfs: scrub |
3452 |
break; |
d9d181c1b Btrfs: rename the... |
3453 |
ret = scrub_pages(sctx, bytenr, BTRFS_SUPER_INFO_SIZE, bytenr, |
a36cf8b89 Btrfs: remove the... |
3454 |
scrub_dev, BTRFS_EXTENT_FLAG_SUPER, gen, i, |
ff023aac3 Btrfs: add code t... |
3455 |
NULL, 1, bytenr); |
a2de733c7 btrfs: scrub |
3456 3457 3458 |
if (ret) return ret; } |
b6bfebc13 Btrfs: cleanup sc... |
3459 |
wait_event(sctx->list_wait, atomic_read(&sctx->bios_in_flight) == 0); |
a2de733c7 btrfs: scrub |
3460 3461 3462 3463 3464 3465 3466 |
return 0; } /* * get a reference count on fs_info->scrub_workers. start worker if necessary */ |
ff023aac3 Btrfs: add code t... |
3467 3468 |
static noinline_for_stack int scrub_workers_get(struct btrfs_fs_info *fs_info, int is_dev_replace) |
a2de733c7 btrfs: scrub |
3469 |
{ |
0dc3b84a7 Btrfs: fix num_wo... |
3470 |
int ret = 0; |
0339ef2f4 btrfs: Replace fs... |
3471 3472 |
int flags = WQ_FREEZABLE | WQ_UNBOUND; int max_active = fs_info->thread_pool_size; |
a2de733c7 btrfs: scrub |
3473 |
|
632dd772f btrfs: reinitiali... |
3474 |
if (fs_info->scrub_workers_refcnt == 0) { |
ff023aac3 Btrfs: add code t... |
3475 |
if (is_dev_replace) |
0339ef2f4 btrfs: Replace fs... |
3476 3477 3478 |
fs_info->scrub_workers = btrfs_alloc_workqueue("btrfs-scrub", flags, 1, 4); |
ff023aac3 Btrfs: add code t... |
3479 |
else |
0339ef2f4 btrfs: Replace fs... |
3480 3481 3482 3483 3484 |
fs_info->scrub_workers = btrfs_alloc_workqueue("btrfs-scrub", flags, max_active, 4); if (!fs_info->scrub_workers) { ret = -ENOMEM; |
0dc3b84a7 Btrfs: fix num_wo... |
3485 |
goto out; |
0339ef2f4 btrfs: Replace fs... |
3486 3487 3488 3489 3490 3491 |
} fs_info->scrub_wr_completion_workers = btrfs_alloc_workqueue("btrfs-scrubwrc", flags, max_active, 2); if (!fs_info->scrub_wr_completion_workers) { ret = -ENOMEM; |
ff023aac3 Btrfs: add code t... |
3492 |
goto out; |
0339ef2f4 btrfs: Replace fs... |
3493 3494 3495 3496 3497 |
} fs_info->scrub_nocow_workers = btrfs_alloc_workqueue("btrfs-scrubnc", flags, 1, 0); if (!fs_info->scrub_nocow_workers) { ret = -ENOMEM; |
ff023aac3 Btrfs: add code t... |
3498 |
goto out; |
0339ef2f4 btrfs: Replace fs... |
3499 |
} |
632dd772f btrfs: reinitiali... |
3500 |
} |
a2de733c7 btrfs: scrub |
3501 |
++fs_info->scrub_workers_refcnt; |
0dc3b84a7 Btrfs: fix num_wo... |
3502 |
out: |
0dc3b84a7 Btrfs: fix num_wo... |
3503 |
return ret; |
a2de733c7 btrfs: scrub |
3504 |
} |
aa1b8cd40 Btrfs: pass fs_in... |
3505 |
static noinline_for_stack void scrub_workers_put(struct btrfs_fs_info *fs_info) |
a2de733c7 btrfs: scrub |
3506 |
{ |
ff023aac3 Btrfs: add code t... |
3507 |
if (--fs_info->scrub_workers_refcnt == 0) { |
0339ef2f4 btrfs: Replace fs... |
3508 3509 3510 |
btrfs_destroy_workqueue(fs_info->scrub_workers); btrfs_destroy_workqueue(fs_info->scrub_wr_completion_workers); btrfs_destroy_workqueue(fs_info->scrub_nocow_workers); |
ff023aac3 Btrfs: add code t... |
3511 |
} |
a2de733c7 btrfs: scrub |
3512 |
WARN_ON(fs_info->scrub_workers_refcnt < 0); |
a2de733c7 btrfs: scrub |
3513 |
} |
aa1b8cd40 Btrfs: pass fs_in... |
3514 3515 |
int btrfs_scrub_dev(struct btrfs_fs_info *fs_info, u64 devid, u64 start, u64 end, struct btrfs_scrub_progress *progress, |
63a212abc Btrfs: disallow s... |
3516 |
int readonly, int is_dev_replace) |
a2de733c7 btrfs: scrub |
3517 |
{ |
d9d181c1b Btrfs: rename the... |
3518 |
struct scrub_ctx *sctx; |
a2de733c7 btrfs: scrub |
3519 3520 |
int ret; struct btrfs_device *dev; |
5d68da3b8 Btrfs: don't writ... |
3521 |
struct rcu_string *name; |
a2de733c7 btrfs: scrub |
3522 |
|
aa1b8cd40 Btrfs: pass fs_in... |
3523 |
if (btrfs_fs_closing(fs_info)) |
a2de733c7 btrfs: scrub |
3524 |
return -EINVAL; |
aa1b8cd40 Btrfs: pass fs_in... |
3525 |
if (fs_info->chunk_root->nodesize > BTRFS_STRIPE_LEN) { |
b5d67f64f Btrfs: change scr... |
3526 3527 3528 3529 3530 |
/* * in this case scrub is unable to calculate the checksum * the way scrub is implemented. Do not handle this * situation at all because it won't ever happen. */ |
efe120a06 Btrfs: convert pr... |
3531 3532 |
btrfs_err(fs_info, "scrub: size assumption nodesize <= BTRFS_STRIPE_LEN (%d <= %d) fails", |
aa1b8cd40 Btrfs: pass fs_in... |
3533 |
fs_info->chunk_root->nodesize, BTRFS_STRIPE_LEN); |
b5d67f64f Btrfs: change scr... |
3534 3535 |
return -EINVAL; } |
aa1b8cd40 Btrfs: pass fs_in... |
3536 |
if (fs_info->chunk_root->sectorsize != PAGE_SIZE) { |
b5d67f64f Btrfs: change scr... |
3537 |
/* not supported for data w/o checksums */ |
efe120a06 Btrfs: convert pr... |
3538 3539 3540 |
btrfs_err(fs_info, "scrub: size assumption sectorsize != PAGE_SIZE " "(%d != %lu) fails", |
27f9f0235 Btrfs: Format PAG... |
3541 |
fs_info->chunk_root->sectorsize, PAGE_SIZE); |
a2de733c7 btrfs: scrub |
3542 3543 |
return -EINVAL; } |
7a9e99876 Btrfs: make the s... |
3544 3545 3546 3547 3548 3549 3550 3551 |
if (fs_info->chunk_root->nodesize > PAGE_SIZE * SCRUB_MAX_PAGES_PER_BLOCK || fs_info->chunk_root->sectorsize > PAGE_SIZE * SCRUB_MAX_PAGES_PER_BLOCK) { /* * would exhaust the array bounds of pagev member in * struct scrub_block */ |
efe120a06 Btrfs: convert pr... |
3552 3553 |
btrfs_err(fs_info, "scrub: size assumption nodesize and sectorsize " "<= SCRUB_MAX_PAGES_PER_BLOCK (%d <= %d && %d <= %d) fails", |
7a9e99876 Btrfs: make the s... |
3554 3555 3556 3557 3558 3559 |
fs_info->chunk_root->nodesize, SCRUB_MAX_PAGES_PER_BLOCK, fs_info->chunk_root->sectorsize, SCRUB_MAX_PAGES_PER_BLOCK); return -EINVAL; } |
a2de733c7 btrfs: scrub |
3560 |
|
aa1b8cd40 Btrfs: pass fs_in... |
3561 3562 |
mutex_lock(&fs_info->fs_devices->device_list_mutex); dev = btrfs_find_device(fs_info, devid, NULL, NULL); |
63a212abc Btrfs: disallow s... |
3563 |
if (!dev || (dev->missing && !is_dev_replace)) { |
aa1b8cd40 Btrfs: pass fs_in... |
3564 |
mutex_unlock(&fs_info->fs_devices->device_list_mutex); |
a2de733c7 btrfs: scrub |
3565 3566 |
return -ENODEV; } |
a2de733c7 btrfs: scrub |
3567 |
|
5d68da3b8 Btrfs: don't writ... |
3568 3569 3570 3571 3572 3573 3574 3575 3576 |
if (!is_dev_replace && !readonly && !dev->writeable) { mutex_unlock(&fs_info->fs_devices->device_list_mutex); rcu_read_lock(); name = rcu_dereference(dev->name); btrfs_err(fs_info, "scrub: device %s is not writable", name->str); rcu_read_unlock(); return -EROFS; } |
3b7a016f4 Btrfs: avoid unne... |
3577 |
mutex_lock(&fs_info->scrub_lock); |
63a212abc Btrfs: disallow s... |
3578 |
if (!dev->in_fs_metadata || dev->is_tgtdev_for_dev_replace) { |
a2de733c7 btrfs: scrub |
3579 |
mutex_unlock(&fs_info->scrub_lock); |
aa1b8cd40 Btrfs: pass fs_in... |
3580 |
mutex_unlock(&fs_info->fs_devices->device_list_mutex); |
aa1b8cd40 Btrfs: pass fs_in... |
3581 |
return -EIO; |
a2de733c7 btrfs: scrub |
3582 |
} |
8dabb7420 Btrfs: change cor... |
3583 3584 3585 3586 3587 |
btrfs_dev_replace_lock(&fs_info->dev_replace); if (dev->scrub_device || (!is_dev_replace && btrfs_dev_replace_is_ongoing(&fs_info->dev_replace))) { btrfs_dev_replace_unlock(&fs_info->dev_replace); |
a2de733c7 btrfs: scrub |
3588 |
mutex_unlock(&fs_info->scrub_lock); |
aa1b8cd40 Btrfs: pass fs_in... |
3589 |
mutex_unlock(&fs_info->fs_devices->device_list_mutex); |
a2de733c7 btrfs: scrub |
3590 3591 |
return -EINPROGRESS; } |
8dabb7420 Btrfs: change cor... |
3592 |
btrfs_dev_replace_unlock(&fs_info->dev_replace); |
3b7a016f4 Btrfs: avoid unne... |
3593 3594 3595 3596 3597 3598 3599 |
ret = scrub_workers_get(fs_info, is_dev_replace); if (ret) { mutex_unlock(&fs_info->scrub_lock); mutex_unlock(&fs_info->fs_devices->device_list_mutex); return ret; } |
63a212abc Btrfs: disallow s... |
3600 |
sctx = scrub_setup_ctx(dev, is_dev_replace); |
d9d181c1b Btrfs: rename the... |
3601 |
if (IS_ERR(sctx)) { |
a2de733c7 btrfs: scrub |
3602 |
mutex_unlock(&fs_info->scrub_lock); |
aa1b8cd40 Btrfs: pass fs_in... |
3603 3604 |
mutex_unlock(&fs_info->fs_devices->device_list_mutex); scrub_workers_put(fs_info); |
d9d181c1b Btrfs: rename the... |
3605 |
return PTR_ERR(sctx); |
a2de733c7 btrfs: scrub |
3606 |
} |
d9d181c1b Btrfs: rename the... |
3607 3608 |
sctx->readonly = readonly; dev->scrub_device = sctx; |
3cb0929ad Btrfs: fix wrong ... |
3609 |
mutex_unlock(&fs_info->fs_devices->device_list_mutex); |
a2de733c7 btrfs: scrub |
3610 |
|
3cb0929ad Btrfs: fix wrong ... |
3611 3612 3613 3614 |
/* * checking @scrub_pause_req here, we can avoid * race between committing transaction and scrubbing. */ |
cb7ab0215 Btrfs: wrap repea... |
3615 |
__scrub_blocked_if_needed(fs_info); |
a2de733c7 btrfs: scrub |
3616 3617 |
atomic_inc(&fs_info->scrubs_running); mutex_unlock(&fs_info->scrub_lock); |
a2de733c7 btrfs: scrub |
3618 |
|
ff023aac3 Btrfs: add code t... |
3619 |
if (!is_dev_replace) { |
9b011adfe Btrfs: remove scr... |
3620 3621 3622 3623 |
/* * by holding device list mutex, we can * kick off writing super in log tree sync. */ |
3cb0929ad Btrfs: fix wrong ... |
3624 |
mutex_lock(&fs_info->fs_devices->device_list_mutex); |
ff023aac3 Btrfs: add code t... |
3625 |
ret = scrub_supers(sctx, dev); |
3cb0929ad Btrfs: fix wrong ... |
3626 |
mutex_unlock(&fs_info->fs_devices->device_list_mutex); |
ff023aac3 Btrfs: add code t... |
3627 |
} |
a2de733c7 btrfs: scrub |
3628 3629 |
if (!ret) |
ff023aac3 Btrfs: add code t... |
3630 3631 |
ret = scrub_enumerate_chunks(sctx, dev, start, end, is_dev_replace); |
a2de733c7 btrfs: scrub |
3632 |
|
b6bfebc13 Btrfs: cleanup sc... |
3633 |
wait_event(sctx->list_wait, atomic_read(&sctx->bios_in_flight) == 0); |
a2de733c7 btrfs: scrub |
3634 3635 |
atomic_dec(&fs_info->scrubs_running); wake_up(&fs_info->scrub_pause_wait); |
b6bfebc13 Btrfs: cleanup sc... |
3636 |
wait_event(sctx->list_wait, atomic_read(&sctx->workers_pending) == 0); |
0ef8e4515 btrfs scrub: add ... |
3637 |
|
a2de733c7 btrfs: scrub |
3638 |
if (progress) |
d9d181c1b Btrfs: rename the... |
3639 |
memcpy(progress, &sctx->stat, sizeof(*progress)); |
a2de733c7 btrfs: scrub |
3640 3641 3642 |
mutex_lock(&fs_info->scrub_lock); dev->scrub_device = NULL; |
3b7a016f4 Btrfs: avoid unne... |
3643 |
scrub_workers_put(fs_info); |
a2de733c7 btrfs: scrub |
3644 |
mutex_unlock(&fs_info->scrub_lock); |
d9d181c1b Btrfs: rename the... |
3645 |
scrub_free_ctx(sctx); |
a2de733c7 btrfs: scrub |
3646 3647 3648 |
return ret; } |
143bede52 btrfs: return voi... |
3649 |
void btrfs_scrub_pause(struct btrfs_root *root) |
a2de733c7 btrfs: scrub |
3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 |
{ struct btrfs_fs_info *fs_info = root->fs_info; mutex_lock(&fs_info->scrub_lock); atomic_inc(&fs_info->scrub_pause_req); while (atomic_read(&fs_info->scrubs_paused) != atomic_read(&fs_info->scrubs_running)) { mutex_unlock(&fs_info->scrub_lock); wait_event(fs_info->scrub_pause_wait, atomic_read(&fs_info->scrubs_paused) == atomic_read(&fs_info->scrubs_running)); mutex_lock(&fs_info->scrub_lock); } mutex_unlock(&fs_info->scrub_lock); |
a2de733c7 btrfs: scrub |
3664 |
} |
143bede52 btrfs: return voi... |
3665 |
void btrfs_scrub_continue(struct btrfs_root *root) |
a2de733c7 btrfs: scrub |
3666 3667 3668 3669 3670 |
{ struct btrfs_fs_info *fs_info = root->fs_info; atomic_dec(&fs_info->scrub_pause_req); wake_up(&fs_info->scrub_pause_wait); |
a2de733c7 btrfs: scrub |
3671 |
} |
aa1b8cd40 Btrfs: pass fs_in... |
3672 |
int btrfs_scrub_cancel(struct btrfs_fs_info *fs_info) |
a2de733c7 btrfs: scrub |
3673 |
{ |
a2de733c7 btrfs: scrub |
3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 |
mutex_lock(&fs_info->scrub_lock); if (!atomic_read(&fs_info->scrubs_running)) { mutex_unlock(&fs_info->scrub_lock); return -ENOTCONN; } atomic_inc(&fs_info->scrub_cancel_req); while (atomic_read(&fs_info->scrubs_running)) { mutex_unlock(&fs_info->scrub_lock); wait_event(fs_info->scrub_pause_wait, atomic_read(&fs_info->scrubs_running) == 0); mutex_lock(&fs_info->scrub_lock); } atomic_dec(&fs_info->scrub_cancel_req); mutex_unlock(&fs_info->scrub_lock); return 0; } |
aa1b8cd40 Btrfs: pass fs_in... |
3692 3693 |
int btrfs_scrub_cancel_dev(struct btrfs_fs_info *fs_info, struct btrfs_device *dev) |
49b25e054 btrfs: enhance tr... |
3694 |
{ |
d9d181c1b Btrfs: rename the... |
3695 |
struct scrub_ctx *sctx; |
a2de733c7 btrfs: scrub |
3696 3697 |
mutex_lock(&fs_info->scrub_lock); |
d9d181c1b Btrfs: rename the... |
3698 3699 |
sctx = dev->scrub_device; if (!sctx) { |
a2de733c7 btrfs: scrub |
3700 3701 3702 |
mutex_unlock(&fs_info->scrub_lock); return -ENOTCONN; } |
d9d181c1b Btrfs: rename the... |
3703 |
atomic_inc(&sctx->cancel_req); |
a2de733c7 btrfs: scrub |
3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 |
while (dev->scrub_device) { mutex_unlock(&fs_info->scrub_lock); wait_event(fs_info->scrub_pause_wait, dev->scrub_device == NULL); mutex_lock(&fs_info->scrub_lock); } mutex_unlock(&fs_info->scrub_lock); return 0; } |
1623edebe Btrfs: minor clea... |
3714 |
|
a2de733c7 btrfs: scrub |
3715 3716 3717 3718 |
int btrfs_scrub_progress(struct btrfs_root *root, u64 devid, struct btrfs_scrub_progress *progress) { struct btrfs_device *dev; |
d9d181c1b Btrfs: rename the... |
3719 |
struct scrub_ctx *sctx = NULL; |
a2de733c7 btrfs: scrub |
3720 3721 |
mutex_lock(&root->fs_info->fs_devices->device_list_mutex); |
aa1b8cd40 Btrfs: pass fs_in... |
3722 |
dev = btrfs_find_device(root->fs_info, devid, NULL, NULL); |
a2de733c7 btrfs: scrub |
3723 |
if (dev) |
d9d181c1b Btrfs: rename the... |
3724 3725 3726 |
sctx = dev->scrub_device; if (sctx) memcpy(progress, &sctx->stat, sizeof(*progress)); |
a2de733c7 btrfs: scrub |
3727 |
mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); |
d9d181c1b Btrfs: rename the... |
3728 |
return dev ? (sctx ? 0 : -ENOTCONN) : -ENODEV; |
a2de733c7 btrfs: scrub |
3729 |
} |
ff023aac3 Btrfs: add code t... |
3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 |
static void scrub_remap_extent(struct btrfs_fs_info *fs_info, u64 extent_logical, u64 extent_len, u64 *extent_physical, struct btrfs_device **extent_dev, int *extent_mirror_num) { u64 mapped_length; struct btrfs_bio *bbio = NULL; int ret; mapped_length = extent_len; ret = btrfs_map_block(fs_info, READ, extent_logical, &mapped_length, &bbio, 0); if (ret || !bbio || mapped_length < extent_len || !bbio->stripes[0].dev->bdev) { kfree(bbio); return; } *extent_physical = bbio->stripes[0].physical; *extent_mirror_num = bbio->mirror_num; *extent_dev = bbio->stripes[0].dev; kfree(bbio); } static int scrub_setup_wr_ctx(struct scrub_ctx *sctx, struct scrub_wr_ctx *wr_ctx, struct btrfs_fs_info *fs_info, struct btrfs_device *dev, int is_dev_replace) { WARN_ON(wr_ctx->wr_curr_bio != NULL); mutex_init(&wr_ctx->wr_lock); wr_ctx->wr_curr_bio = NULL; if (!is_dev_replace) return 0; WARN_ON(!dev->bdev); wr_ctx->pages_per_wr_bio = min_t(int, SCRUB_PAGES_PER_WR_BIO, bio_get_nr_vecs(dev->bdev)); wr_ctx->tgtdev = dev; atomic_set(&wr_ctx->flush_all_writes, 0); return 0; } static void scrub_free_wr_ctx(struct scrub_wr_ctx *wr_ctx) { mutex_lock(&wr_ctx->wr_lock); kfree(wr_ctx->wr_curr_bio); wr_ctx->wr_curr_bio = NULL; mutex_unlock(&wr_ctx->wr_lock); } static int copy_nocow_pages(struct scrub_ctx *sctx, u64 logical, u64 len, int mirror_num, u64 physical_for_dev_replace) { struct scrub_copy_nocow_ctx *nocow_ctx; struct btrfs_fs_info *fs_info = sctx->dev_root->fs_info; nocow_ctx = kzalloc(sizeof(*nocow_ctx), GFP_NOFS); if (!nocow_ctx) { spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); return -ENOMEM; } scrub_pending_trans_workers_inc(sctx); nocow_ctx->sctx = sctx; nocow_ctx->logical = logical; nocow_ctx->len = len; nocow_ctx->mirror_num = mirror_num; nocow_ctx->physical_for_dev_replace = physical_for_dev_replace; |
9e0af2376 Btrfs: fix task h... |
3806 3807 |
btrfs_init_work(&nocow_ctx->work, btrfs_scrubnc_helper, copy_nocow_pages_worker, NULL, NULL); |
652f25a29 Btrfs: improve re... |
3808 |
INIT_LIST_HEAD(&nocow_ctx->inodes); |
0339ef2f4 btrfs: Replace fs... |
3809 3810 |
btrfs_queue_work(fs_info->scrub_nocow_workers, &nocow_ctx->work); |
ff023aac3 Btrfs: add code t... |
3811 3812 3813 |
return 0; } |
652f25a29 Btrfs: improve re... |
3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 |
static int record_inode_for_nocow(u64 inum, u64 offset, u64 root, void *ctx) { struct scrub_copy_nocow_ctx *nocow_ctx = ctx; struct scrub_nocow_inode *nocow_inode; nocow_inode = kzalloc(sizeof(*nocow_inode), GFP_NOFS); if (!nocow_inode) return -ENOMEM; nocow_inode->inum = inum; nocow_inode->offset = offset; nocow_inode->root = root; list_add_tail(&nocow_inode->list, &nocow_ctx->inodes); return 0; } #define COPY_COMPLETE 1 |
ff023aac3 Btrfs: add code t... |
3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 |
static void copy_nocow_pages_worker(struct btrfs_work *work) { struct scrub_copy_nocow_ctx *nocow_ctx = container_of(work, struct scrub_copy_nocow_ctx, work); struct scrub_ctx *sctx = nocow_ctx->sctx; u64 logical = nocow_ctx->logical; u64 len = nocow_ctx->len; int mirror_num = nocow_ctx->mirror_num; u64 physical_for_dev_replace = nocow_ctx->physical_for_dev_replace; int ret; struct btrfs_trans_handle *trans = NULL; struct btrfs_fs_info *fs_info; struct btrfs_path *path; struct btrfs_root *root; int not_written = 0; fs_info = sctx->dev_root->fs_info; root = fs_info->extent_root; path = btrfs_alloc_path(); if (!path) { spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); not_written = 1; goto out; } trans = btrfs_join_transaction(root); if (IS_ERR(trans)) { not_written = 1; goto out; } ret = iterate_inodes_from_logical(logical, fs_info, path, |
652f25a29 Btrfs: improve re... |
3865 |
record_inode_for_nocow, nocow_ctx); |
ff023aac3 Btrfs: add code t... |
3866 |
if (ret != 0 && ret != -ENOENT) { |
efe120a06 Btrfs: convert pr... |
3867 3868 |
btrfs_warn(fs_info, "iterate_inodes_from_logical() failed: log %llu, " "phys %llu, len %llu, mir %u, ret %d", |
118a0a251 Btrfs: Format mir... |
3869 3870 |
logical, physical_for_dev_replace, len, mirror_num, ret); |
ff023aac3 Btrfs: add code t... |
3871 3872 3873 |
not_written = 1; goto out; } |
652f25a29 Btrfs: improve re... |
3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 |
btrfs_end_transaction(trans, root); trans = NULL; while (!list_empty(&nocow_ctx->inodes)) { struct scrub_nocow_inode *entry; entry = list_first_entry(&nocow_ctx->inodes, struct scrub_nocow_inode, list); list_del_init(&entry->list); ret = copy_nocow_pages_for_inode(entry->inum, entry->offset, entry->root, nocow_ctx); kfree(entry); if (ret == COPY_COMPLETE) { ret = 0; break; } else if (ret) { break; } } |
ff023aac3 Btrfs: add code t... |
3892 |
out: |
652f25a29 Btrfs: improve re... |
3893 3894 3895 3896 3897 3898 3899 3900 |
while (!list_empty(&nocow_ctx->inodes)) { struct scrub_nocow_inode *entry; entry = list_first_entry(&nocow_ctx->inodes, struct scrub_nocow_inode, list); list_del_init(&entry->list); kfree(entry); } |
ff023aac3 Btrfs: add code t... |
3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 |
if (trans && !IS_ERR(trans)) btrfs_end_transaction(trans, root); if (not_written) btrfs_dev_replace_stats_inc(&fs_info->dev_replace. num_uncorrectable_read_errors); btrfs_free_path(path); kfree(nocow_ctx); scrub_pending_trans_workers_dec(sctx); } |
321592427 btrfs: fix dead l... |
3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 |
static int check_extent_to_block(struct inode *inode, u64 start, u64 len, u64 logical) { struct extent_state *cached_state = NULL; struct btrfs_ordered_extent *ordered; struct extent_io_tree *io_tree; struct extent_map *em; u64 lockstart = start, lockend = start + len - 1; int ret = 0; io_tree = &BTRFS_I(inode)->io_tree; lock_extent_bits(io_tree, lockstart, lockend, 0, &cached_state); ordered = btrfs_lookup_ordered_range(inode, lockstart, len); if (ordered) { btrfs_put_ordered_extent(ordered); ret = 1; goto out_unlock; } em = btrfs_get_extent(inode, NULL, 0, start, len, 0); if (IS_ERR(em)) { ret = PTR_ERR(em); goto out_unlock; } /* * This extent does not actually cover the logical extent anymore, * move on to the next inode. */ if (em->block_start > logical || em->block_start + em->block_len < logical + len) { free_extent_map(em); ret = 1; goto out_unlock; } free_extent_map(em); out_unlock: unlock_extent_cached(io_tree, lockstart, lockend, &cached_state, GFP_NOFS); return ret; } |
652f25a29 Btrfs: improve re... |
3955 3956 |
static int copy_nocow_pages_for_inode(u64 inum, u64 offset, u64 root, struct scrub_copy_nocow_ctx *nocow_ctx) |
ff023aac3 Btrfs: add code t... |
3957 |
{ |
826aa0a82 Btrfs: cleanup th... |
3958 |
struct btrfs_fs_info *fs_info = nocow_ctx->sctx->dev_root->fs_info; |
ff023aac3 Btrfs: add code t... |
3959 |
struct btrfs_key key; |
826aa0a82 Btrfs: cleanup th... |
3960 3961 |
struct inode *inode; struct page *page; |
ff023aac3 Btrfs: add code t... |
3962 |
struct btrfs_root *local_root; |
652f25a29 Btrfs: improve re... |
3963 |
struct extent_io_tree *io_tree; |
ff023aac3 Btrfs: add code t... |
3964 |
u64 physical_for_dev_replace; |
321592427 btrfs: fix dead l... |
3965 |
u64 nocow_ctx_logical; |
652f25a29 Btrfs: improve re... |
3966 |
u64 len = nocow_ctx->len; |
826aa0a82 Btrfs: cleanup th... |
3967 |
unsigned long index; |
6f1c36055 Btrfs: fix race b... |
3968 |
int srcu_index; |
652f25a29 Btrfs: improve re... |
3969 3970 |
int ret = 0; int err = 0; |
ff023aac3 Btrfs: add code t... |
3971 3972 3973 3974 |
key.objectid = root; key.type = BTRFS_ROOT_ITEM_KEY; key.offset = (u64)-1; |
6f1c36055 Btrfs: fix race b... |
3975 3976 |
srcu_index = srcu_read_lock(&fs_info->subvol_srcu); |
ff023aac3 Btrfs: add code t... |
3977 |
local_root = btrfs_read_fs_root_no_name(fs_info, &key); |
6f1c36055 Btrfs: fix race b... |
3978 3979 |
if (IS_ERR(local_root)) { srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); |
ff023aac3 Btrfs: add code t... |
3980 |
return PTR_ERR(local_root); |
6f1c36055 Btrfs: fix race b... |
3981 |
} |
ff023aac3 Btrfs: add code t... |
3982 3983 3984 3985 3986 |
key.type = BTRFS_INODE_ITEM_KEY; key.objectid = inum; key.offset = 0; inode = btrfs_iget(fs_info->sb, &key, local_root, NULL); |
6f1c36055 Btrfs: fix race b... |
3987 |
srcu_read_unlock(&fs_info->subvol_srcu, srcu_index); |
ff023aac3 Btrfs: add code t... |
3988 3989 |
if (IS_ERR(inode)) return PTR_ERR(inode); |
edd1400be Btrfs: fix severa... |
3990 3991 3992 |
/* Avoid truncate/dio/punch hole.. */ mutex_lock(&inode->i_mutex); inode_dio_wait(inode); |
ff023aac3 Btrfs: add code t... |
3993 |
physical_for_dev_replace = nocow_ctx->physical_for_dev_replace; |
652f25a29 Btrfs: improve re... |
3994 |
io_tree = &BTRFS_I(inode)->io_tree; |
321592427 btrfs: fix dead l... |
3995 |
nocow_ctx_logical = nocow_ctx->logical; |
652f25a29 Btrfs: improve re... |
3996 |
|
321592427 btrfs: fix dead l... |
3997 3998 3999 4000 |
ret = check_extent_to_block(inode, offset, len, nocow_ctx_logical); if (ret) { ret = ret > 0 ? 0 : ret; goto out; |
652f25a29 Btrfs: improve re... |
4001 |
} |
652f25a29 Btrfs: improve re... |
4002 |
|
ff023aac3 Btrfs: add code t... |
4003 |
while (len >= PAGE_CACHE_SIZE) { |
ff023aac3 Btrfs: add code t... |
4004 |
index = offset >> PAGE_CACHE_SHIFT; |
edd1400be Btrfs: fix severa... |
4005 |
again: |
ff023aac3 Btrfs: add code t... |
4006 4007 |
page = find_or_create_page(inode->i_mapping, index, GFP_NOFS); if (!page) { |
efe120a06 Btrfs: convert pr... |
4008 |
btrfs_err(fs_info, "find_or_create_page() failed"); |
ff023aac3 Btrfs: add code t... |
4009 |
ret = -ENOMEM; |
826aa0a82 Btrfs: cleanup th... |
4010 |
goto out; |
ff023aac3 Btrfs: add code t... |
4011 4012 4013 4014 4015 4016 4017 |
} if (PageUptodate(page)) { if (PageDirty(page)) goto next_page; } else { ClearPageError(page); |
321592427 btrfs: fix dead l... |
4018 |
err = extent_read_full_page(io_tree, page, |
652f25a29 Btrfs: improve re... |
4019 4020 |
btrfs_get_extent, nocow_ctx->mirror_num); |
826aa0a82 Btrfs: cleanup th... |
4021 4022 |
if (err) { ret = err; |
ff023aac3 Btrfs: add code t... |
4023 4024 |
goto next_page; } |
edd1400be Btrfs: fix severa... |
4025 |
|
26b258919 Btrfs: fix oops w... |
4026 |
lock_page(page); |
edd1400be Btrfs: fix severa... |
4027 4028 4029 4030 4031 4032 4033 |
/* * If the page has been remove from the page cache, * the data on it is meaningless, because it may be * old one, the new data may be written into the new * page in the page cache. */ if (page->mapping != inode->i_mapping) { |
652f25a29 Btrfs: improve re... |
4034 |
unlock_page(page); |
edd1400be Btrfs: fix severa... |
4035 4036 4037 |
page_cache_release(page); goto again; } |
ff023aac3 Btrfs: add code t... |
4038 4039 4040 4041 4042 |
if (!PageUptodate(page)) { ret = -EIO; goto next_page; } } |
321592427 btrfs: fix dead l... |
4043 4044 4045 4046 4047 4048 4049 |
ret = check_extent_to_block(inode, offset, len, nocow_ctx_logical); if (ret) { ret = ret > 0 ? 0 : ret; goto next_page; } |
826aa0a82 Btrfs: cleanup th... |
4050 4051 4052 4053 |
err = write_page_nocow(nocow_ctx->sctx, physical_for_dev_replace, page); if (err) ret = err; |
ff023aac3 Btrfs: add code t... |
4054 |
next_page: |
826aa0a82 Btrfs: cleanup th... |
4055 4056 4057 4058 4059 |
unlock_page(page); page_cache_release(page); if (ret) break; |
ff023aac3 Btrfs: add code t... |
4060 4061 |
offset += PAGE_CACHE_SIZE; physical_for_dev_replace += PAGE_CACHE_SIZE; |
321592427 btrfs: fix dead l... |
4062 |
nocow_ctx_logical += PAGE_CACHE_SIZE; |
ff023aac3 Btrfs: add code t... |
4063 4064 |
len -= PAGE_CACHE_SIZE; } |
652f25a29 Btrfs: improve re... |
4065 |
ret = COPY_COMPLETE; |
826aa0a82 Btrfs: cleanup th... |
4066 |
out: |
edd1400be Btrfs: fix severa... |
4067 |
mutex_unlock(&inode->i_mutex); |
826aa0a82 Btrfs: cleanup th... |
4068 |
iput(inode); |
ff023aac3 Btrfs: add code t... |
4069 4070 4071 4072 4073 4074 4075 4076 4077 |
return ret; } static int write_page_nocow(struct scrub_ctx *sctx, u64 physical_for_dev_replace, struct page *page) { struct bio *bio; struct btrfs_device *dev; int ret; |
ff023aac3 Btrfs: add code t... |
4078 4079 4080 4081 4082 4083 |
dev = sctx->wr_ctx.tgtdev; if (!dev) return -EIO; if (!dev->bdev) { printk_ratelimited(KERN_WARNING |
efe120a06 Btrfs: convert pr... |
4084 4085 |
"BTRFS: scrub write_page_nocow(bdev == NULL) is unexpected! "); |
ff023aac3 Btrfs: add code t... |
4086 4087 |
return -EIO; } |
9be3395bc Btrfs: use a btrf... |
4088 |
bio = btrfs_io_bio_alloc(GFP_NOFS, 1); |
ff023aac3 Btrfs: add code t... |
4089 4090 4091 4092 4093 4094 |
if (!bio) { spin_lock(&sctx->stat_lock); sctx->stat.malloc_errors++; spin_unlock(&sctx->stat_lock); return -ENOMEM; } |
4f024f379 block: Abstract o... |
4095 4096 |
bio->bi_iter.bi_size = 0; bio->bi_iter.bi_sector = physical_for_dev_replace >> 9; |
ff023aac3 Btrfs: add code t... |
4097 4098 4099 4100 4101 4102 4103 4104 |
bio->bi_bdev = dev->bdev; ret = bio_add_page(bio, page, PAGE_CACHE_SIZE, 0); if (ret != PAGE_CACHE_SIZE) { leave_with_eio: bio_put(bio); btrfs_dev_stat_inc_and_print(dev, BTRFS_DEV_STAT_WRITE_ERRS); return -EIO; } |
ff023aac3 Btrfs: add code t... |
4105 |
|
33879d451 block: submit_bio... |
4106 |
if (btrfsic_submit_bio_wait(WRITE_SYNC, bio)) |
ff023aac3 Btrfs: add code t... |
4107 4108 4109 4110 4111 |
goto leave_with_eio; bio_put(bio); return 0; } |